Online Supplement

Home , Byssinosis

BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s) Thorax

Online Supplement:

Manuscript Title: Nocturnal Hypoxaemia in Interstitial Lung Disease: A Systematic Review

Authors: Yet H Khor, Yvonne Ng, Duncan Sweeney, Christopher J Ryerson

Khor YH, et al. Thorax 2021;0:1–9. doi: 10.1136/thoraxjnl-2020-216749 BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s) Thorax

Table of content:

S1. Database search strategy …………………………………………………………………………………………………. 3

Table S2. Risk of bias assessment tool for non-intervention studies ………………………………………………… 6

S3. Additional reports of included studies ……………………………………………………………………………………….. 8

Table S4. Characteristics of included studies …………………………………………………………………………………… 9

Table S5. Summary of the risk of bias assessment for non-intervention studies ……………………………. 16

Table S6. Summary of the risk of bias assessment for intervention studies ..….…………………………...….. 17 Figure S1. Sensitivity analyses of pooled proportions of clinically significant nocturnal hypoxaemia: a) risk of bias, b) types of study design, and c) study participant numbers ………………………………………… 18

Table S7. Associations of nocturnal hypoxaemia in ILD ……….…………………………………………………………. 21

Table S8. Impacts of supplemental oxygen therapy for nocturnal hypoxaemia in ILD .…………………... 28 Table S9. Impacts of nocturnal hypoxaemia on health-related quality of life and symptoms in ILD .. 29

S1. Database search strategy

Ovid MEDLINE Row Term Hits 1 Nocturnal*.mp. 37042 2 Night*.mp. 85467 3 Sleep*.mp. 204945 4 or/1-3 282024 5 exp Sleep Apnea Syndromes/ 35201 6 (sleep* adj3 (apnea* or apnoea*)).mp. 45550 7 (hypopnea* or hypopnoea*).mp. 11854 8 OSA.mp. 13900 9 SHS.mp. 2782 10 OSAHS.mp. 1439 11 or/5-10 50837 12 4 or 11 286568 13 exp Lung Diseases, Interstitial/ 55448 14 exp Pulmonary Fibrosis/ 23255 15 (Interstitial* adj3 (lung* or pneumon* or pulmon* or fibros*)).mp. 34717 16 ((pulmonary* or lung* or respirat*) adj3 (fibros* or fibrot*)).mp. 35558 17 (diffuse* adj3 (lung* or pneumon* or pulmon* or parenchym*)).mp. 6646 18 ILD.mp. 4097 19 Alveolitis.mp. 6138 20 (pulmonary* adj3 sarcoid*).mp. 5135 21 (asbestosis or silicosis or siderosis or byssinosis or berylliosis or 16667 anthracosilicosis or sillicotuberculosis).mp. 22 Pneumo#onio*.mp. 8180 23 exp Pneumoconiosis/ 20117 24 alveolitis, extrinsic allergic/ or bird fancier's lung/ or farmer's lung/ or silo 4248 filler's disease/ or trichosporonosis/ 25 ((hypersensitiv* or allerg* or lymphocyt* or granulomat*) adj3 (lung* or 25290 pneumon* or pulmon* or respirat* or fibros* or alveolit*)).mp. 26 or/13-25 132436 27 Granulomat*.mp. 40568 28 exp Histiocytosis/ or histiocytosis*.mp. 22376 29 exp Connective tissue diseases/ 302631 30 (connective adj3 tissue adj3 (disorder* or disease*)).mp. 18880 31 exp Rheumatic diseases/ 224776 32 rheumatic*.mp. 65811 33 exp Scleroderma, Systemic/ 20423 34 (scleroderma or sclerosis or polymyositis or dermatomyositis or myositis or 335461 systemic lupus erythematos* or rheumatoid arthritis).mp. 35 or/27-34 658313 36 35 and (lung* or pneumon* or pulmon* or respirat* or fibros* or 59415 alveolit*).mp. 37 26 or 36 174990 38 12 and 37 1462

Embase Row Term Hits 1 Nocturnal*.mp. 57961 2 Night*.mp. 145247 3 Sleep*.mp. 338228 4 or/1-3 461800 5 exp Sleep Apnea Syndromes/ 50228 6 (sleep* adj3 (apnea* or apnoea*)).mp. 72002 7 (hypopnea* or hypopnoea*).mp. 23907 8 OSA.mp. 27698 9 SHS.mp. 3720 10 OSAHS.mp. 2032 11 or/5-10 92677 12 4 or 11 468471 13 exp interstitial lung disease/ 85235 14 exp lung fibrosis/ 84112 15 (Interstitial* adj3 (lung* or pneumon* or pulmon* or fibros*)).mp. 64513 16 ((pulmonary* or lung* or respirat*) adj3 (fibros* or fibrot*)).mp. 62177 17 (diffuse* adj3 (lung* or pneumon* or pulmon* or parenchym*)).mp. 11050 18 ILD.mp. 9536 19 alveolitis.mp. 30855 20 (pulmonary* adj3 sarcoid*).mp. 4973 21 (asbestosis or silicosis or siderosis or byssinosis or berylliosis or 22979 anthracosilicosis or sillicotuberculosis).mp. 22 Pneumo#onio*.mp. 11016 23 exp Pneumoconiosis/ 9619 24 allergic pneumonitis/ or bird breeder lung/ or farmer lung/ or loeffler 12132 pneumonia/ or pigeon breeder lung/ 25 ((hypersensitiv* or allerg* or lymphocyt* or granulomat*) adj3 (lung* or 37088 pneumon* or pulmon* or respirat* or fibros* or alveolit*)).mp. 26 or/13-25 204804 27 Granulomat*.mp. 69785 28 exp Histiocytosis/ or histiocytosis*.mp. 23036 29 exp Connective tissue diseases/ 479974 30 (connective adj3 tissue adj3 (disorder* or disease*)).mp. 33601 31 exp Rheumatic diseases/ 275495 32 Rheumatic$.mp. 115585 33 exp Systemic Sclerosis/ 31571 34 (scleroderma or sclerosis or polymyositis or dermatomyositis or myositis or 627294 systemic lupus erythematos* or rheumatoid arthritis).mp. 35 or/27-34 904182 36 35 and (lung* or pneumon* or pulmon* or respirat* or fibros* or 132791 alveolit*).mp. 37 26 or 36 302636 38 12 and 37 5456

CENTRAL (Cochrane Register of Studies Online) Row Term Hits #1 Nocturnal* 6468 #2 Night* 19353 #3 Sleep* 37407 #4 MeSH descriptor: [Sleep] explode all trees 5528 #5 {or #1-#4} 51499 #6 MeSH descriptor: [Sleep Apnea Syndromes] explode all trees 2541 #7 MeSH descriptor: [Sleep Apnea, Obstructive] explode all trees 1872 #8 sleep* near/3 (apnea* or apnoea*) 6450 #9 hypopnea* or hypopnoea* 2494 #10 OSA 3545 #11 SHS 351 #12 OSAHS 208 #13 {or #6-#12} 7413 #14 #5 or #13 52047 #15 MeSH descriptor: [Lung Diseases, Interstitial] explode all trees 786 #16 MeSH descriptor: [Pulmonary Fibrosis] explode all trees 504 #17 Interstitial* near/3 (lung* or pneumon* or pulmon* or fibros*) 1926 #18 (pulmonary* or lung* or respirat*) near/3 (fibros* or fibrot*) 2174 #19 diffuse* near/3 (lung* or pneumon* or pulmon* or parenchym*) 157 #20 ILD 472 #21 Alveolitis 699 #22 pulmonary* near/3 sarcoid* 202 #23 asbestosis or silicosis or siderosis or byssinosis or berylliosis or 147 anthracosilicosis or sillicotuberculosis #24 Pneumo?onio* 71 #25 MeSH descriptor: [Pneumoconiosis] explode all trees 97 #26 MeSH descriptor: [Alveolitis, Extrinsic Allergic] explode all trees 26 #27 (bird* or farmer* or pigeon* or avian* or budgerigar*) near/3 (lung* or 108 disease*) #28 (hypersensitiv* or allerg* or lymphocyt* or granulomat*) near/3 (lung* or 2367 pneumon* or pulmon* or respirat* or fibros* or alveolit*) #29 MeSH descriptor: [Histiocytosis, Langerhans-Cell] explode all trees 19 #30 {or #15-#29} 6648 #31 Granulomat* 595 #32 MeSH descriptor: [Connective Tissue Diseases] explode all trees 9395 #33 (connective near/3 tissue near/3 (disorder* or disease*)) 2239 #34 MeSH descriptor: [Rheumatic Diseases] explode all trees 15198 #35 rheumatic* 7433 #36 MeSH descriptor: [Scleroderma, Systemic] explode all trees 532 #37 scleroderma* or sclerosis* or polymyositis or dermatomyositis or myositis or 30000 systemic lupus erythematos* or rheumatoid arthritis #38 {or #31-#37} 45764 #39 lung* or pneumon* or pulmon* or respirat* or fibros* or alveolit* 168642 #40 #38 and #39 3714 #41 #30 or #40 9763 #42 #14 and #41 410

Table S2. Risk of bias assessment for non-intervention studies*

Risk of bias items Risk of bias levels (Select HIGH RISK if information is unclear or Points not provided) scored 1. Was the study target Yes (LOW RISK): The study’s target population was a close 0 population a close representation of the general ILD/IPF population. representation of the No (HIGH RISK): The study’s target population was clearly NOT 1 general ILD/IPF representative of the general ILD/IPF population. population in relation to relevant variables, e.g. age, sex, etc? 2. Was the sampling frame Yes (LOW RISK): The sampling frame was a true or close 0 a true or close representation of the target population. representation of the No (HIGH RISK): The sampling frame was NOT a true or close 1 target population? representation of the target population. 3. Was some form of Yes (LOW RISK): A consecutive sampling was undertaken, OR, 0 random selection used to some form of random selection was used to select the sample select the sample, OR, (e.g. simple random sampling, stratified random sampling, was a consecutive cluster sampling, systematic sampling)? sampling undertaken? No (HIGH RISK): A census was NOT undertaken, AND some form 1 of random selection was NOT used to select the sample. 4. Was the likelihood of Yes (LOW RISK): The participation rate for the study was ≥75%, 0 non-participation bias OR, an analysis was performed that showed no significant minimal? difference in relevant demographic characteristics between participants and non-participants No (HIGH RISK): The participation rate was <75%, and if any 1 analysis comparing responders and non-responders was done, it showed a significant difference in relevant demographic characteristics between participants and non-participants 5. Were data collected Yes (LOW RISK): All data were collected directly from the 0 directly from the subjects subjects. (as opposed to medical No (HIGH RISK): In some instances, data were collected from 1 records or proxy)? medical records/proxy. 6. Were acceptable case Yes (LOW RISK): An acceptable case definition was used 0 definition of ILD used? No (HIGH RISK): An acceptable case definition was NOT used 1 7. Were acceptable case Yes (LOW RISK): An acceptable case definition was used 0 definition of nocturnal No (HIGH RISK): An acceptable case definition was NOT used 1 hypoxaemia and/or desaturation used? 8. Was the study Yes (LOW RISK): The study instrument had been shown to have 0 instrument that reliability and validity (if this was necessary). measured nocturnal No (HIGH RISK): The study instrument had NOT been shown to 1 hypoxaemia and/or have reliability or validity (if this was necessary) desaturation reliable and valid (if necessary)? 9. Was the same mode of Yes (LOW RISK): The same mode of data collection was used for 0 data collection used for all subjects (assume so unless specified) all subjects? No (HIGH RISK): The same mode of data collection was NOT used 1 for all subjects 10. Was the length of the Yes (LOW RISK): The shortest prevalence period for nocturnal 0 shortest prevalence hypoxaemia/desaturation was appropriate period for nocturnal No (HIGH RISK): The shortest prevalence period for nocturnal 1 hypoxaemia/ hypoxaemia/desaturation was not appropriate desaturation appropriate (*If not applicable (e.g.

cross-sectional design), select 0) 11. Were the numerator(s) Yes (LOW RISK): The paper presented appropriate numerator(s) 0 and denominator(s) for AND denominator(s) for the parameter of interest the parameter of interest No (HIGH RISK): The paper did present numerator(s) AND 1 appropriate? denominator(s) for the parameter of interest but one or more of these were inappropriate. Summary LOW RISK: Further research is very unlikely to change the 1 confidence in the estimate MODERATE RISK: Further research is likely to have an important 2 impact on the confidence in the estimate and may change the estimate HIGH RISK: Further research is very likely to have an important 3 impact on the confidence in the estimate and is likely to change the estimate.

* Adapted from Hoy et al. Assessing risk of bias in prevalence studies: modification of an existing tool and evidence of interrater agreement. J Clin Epidemiol 2012: 65(9): 934-939.

S3. Additional reports of included studies

Data from additional reports were reviewed and collated, in addition to the primary studies.

Al Aghbari J, et al. Supplemental oxygen improves sleep disordered breathing but not quality of life in patients with interstitial lung disease. Respirology. 2016;21:147.  Troy L, et al. Supplemental oxygen improves sleep disordered breathing in patients with interstitial lung disease. Respirology 2015;20:58.  Troy L, et al. Supplemental Oxygen Improves Sleep Disordered Breathing but Not Quality of Life in Patients with Interstitial Lung Disease. Am J Respir Crit Care Med. 2016;193:A6520.  Troy LK, et al. Oxygen attenuates oxidative stress in ILD patients with nocturnal and exercise- induced hypoxaemia." Respirology 2017;22:32.

Bosi M, et al. OSA and Prolonged Oxygen Desaturation During Sleep are Strong Predictors of Poor Outcome in IPF." Lung 2017;195(5):643-651.  Bosi M, et al. Quality of life in idiopathic pulmonary fibrosis: The impact of sleep disordered breathing. Respir Med. 2019;147:51-57.

Corte TJ, et al. Elevated nocturnal desaturation index predicts mortality in interstitial lung disease. Sarcoidosis Vasc Diffuse Lung Dis. 2012;29(1):41-50.  Corte TJ, et al. Elevated nocturnal oxygen desaturation index predicts higher short-term mortality in interstitial lung disease. Thorax. 2009;64:A5.  Corte TJ, et al. Nocturnal oxygen desaturation is associated with increased mortality in interstitial lung disease patients." Am J Respir Crit Care Med. 2010;181.

Lee RNC, et al. Disordered breathing during sleep and exercise in idiopathic pulmonary fibrosis and the role of biomarkers. QJM 2015;108(4):315-323.  Lee RNC, et al. Comparison of sleep and exercise gas exchange in patients with stable idiopathic pulmonary fibrosis." Am J Respir Crit Care Med. 2010;181.  Lee R, et al. Gas exchange during sleep and exercise in patients with idiopathic pulmonary fibrosis (IPF)." Irish Journal of Medical Science 2012;181:S415-S416.  Lee RNC, et al. Ventilation and gas exchange during sleep and exercise in patients with idiopathic pulmonary fibrosis: Physiological and molecular findings. Am J Respir Crit Care Med. 2013;187(MeetingAbstracts).

Troy LK, et al. Nocturnal hypoxaemia is associated with adverse outcomes in interstitial lung disease." Respirology 2019;24(10):996-1004.  Troy L, et al. Sleep-disordered breathing in patients with interstitial lung disease. Respirology 2012;17:75.

Table S4. Characteristics of included studies

Reference Location Funding Study design Participant ILD subtypes; Inclusion Diagnostic Clinical Oxygen use/ Key exclusion criteria number diagnostic of OSA tool for characteristicsa daytime criteria nocturnal hypoxaemia hypoxaemia Ahmed Egypt Nil Prospective 20 IPF; 2011 Yes In-laboratory 30% male Daytime resting Smoking history; 2018 cross-sectional international PSG Age: 47.1±11.8 SpO2: occupational exposure; study, single guideline 94.89±2.47% chest trauma or infection centre in the last 4 weeks; cardiovascular/liver/renal diseases Al Aghbari Australia Unclear RCT, single 12 ILD; not provided Yes In-laboratory 42% male Not provided Not provided 2016b centre PSG Age: 66.5±8.8 FVC: 68±15.3 DLCO: 48.5±11.7 Avdeev Russia Unclear Prospective 25 IPF; not provided Yes Unclear Age: 57±8 Not provided Not provided 2011b cross-sectional FVC: 80±18 study, number of DLCO: 38±15 study centre unclear Aydogdu Turkey Unclear Prospective 37 ILD; not Yes PSG, study Not provided Not provided Not provided 2006b,c cross-sectional specified location study, number of unclear study centre unclear Bingol Turkey Unclear Prospective 29 Sarcoidosis; Yes In-laboratory 7% male Daytime resting ICU admission or 2015 cross-sectional histological PSG Age: 46.4±11.7 PaO2 (Stages II- treatment modifications in study, single confirmation FVC: 97±22 III sarcoidosis): the last 3 months; centre DLCO: 90±24 82.3±8.7mmHg neurosarcoidosis; severe psychiatric disorder; LRTI within last month; non- ILD chronic lung diseases; high risk for OSA (upper respiratory tract pathologies, BMI ≥30); chronic use of medications that can affect sleep architecture) Bosi 2017 Italy Nil Prospective 35 IPF; 2011 Yes Home PSG 77% male Patients using Alcohol consumption; cohort study, international Age: 68.7±9.2 oxygen therapy insomnia; use of single centre guideline FVC: 72.2±19.6 during benzodiazepines, other DLCO: 45.6±15.3 wakefulness neurological or and/or sleep psychiatric drugs, or beta were excluded blockers; neurological or psychiatric disorders; non-IPF chronic lung diseases; serious and

unstable chronic diseases; Bye 1984 Australia Public Prospective 13 ILD; CXR Yes In-laboratory 85% male Daytime resting Not provided cross-sectional changes ± PSG Age: 51.8±13.9 PaO2: study, number of histological FVC: 66±16 74±11mmHg study centre confirmation DLCO: 62±21 unclear Canora Italy Nil Prospective 100 IPF; 2011 Yes Home PSG IPF IPF Clinically unstable 2018 cross-sectional international 74% male Daytime resting disease study, single guideline Age: 69.9±7.1 PaO2: centre Other ILD: not FVC: 71.5±23.2 70±11mmHg specified DLCO: 47.5±19.4 Non-IPF Non-IPF 63% male Daytime resting Age: 66.8±8.8 PaO2: FVC: 67.4±20.8 69±13mmHg DLCO: 50±18.7 Cardoso Portugal Unclear Prospective 49 ILD; not Yes Home PSG 53% male Patients with High risk for sleep 2018 cross-sectional specified Age: 67.2±12.2 daytime resting disorders (BMI ≥30, study, single FVC: 86.1±18.9 PaO2 <60mmHg cranial-facial deformities, centre DLCO: 65.8±18.6 were excluded and benzodiazepine use); neuromuscular diseases; severe psychiatric diseases; non-ILD chronic lung diseases; previously diagnosed OSA Chudiwal India Unclear Retrospective 15 ILD; HRCT Yes Home PSG Age: 63.4±7.5 Not provided Not provided 2012b cross-sectional confirmation study, number of study centre unclear Clark 2001 UK Public Prospective 50 IPF; clinical and Yes Overnight 66% male 2 patients using Not provided cross-sectional lung function oximetry Age: 67.7±8.7 nocturnal study, single assessment with FVC: 80.9±25 oxygen therapy centre histological ± HRCT confirmation Corte UK Industry Retrospective 134 ILD; not Yes Overnight 56% male 17% with Corresponding 2012 cohort study, specified oximetry Age: 59±14 daytime resting pulmonary function and single centre FVC: 65.5±21.3 SpO2 <92%; echocardiography not DLCO: 37.4±16.1 47% with 6MWT available end SpO2 ≤88% Desideri Italy Unclear Prospective 15 ILD with UIP Yes Overnight 73% male Not provided Respiratory failure 2013b cross-sectional pattern; not cardio- Age: 70.9±6.8 study, single specified respiratory centre polygraphy

Ernst Argentina Unclear Prospective 17 IPF; 2013 Yes Overnight 65% male Not provided Not provided 2015b cross-sectional international cardio- Age: 66±2.4 study, number of guideline respiratory study centre polygraphy unclear Frija- France Unclear Retrospective 33 IPF; not Yes PSG, study 70% male None with Not provided Masson cross-sectional specified location Age: 59.8±11.6 oxygen 2019b study, number of unclear FVC: 61±23 requirement study centre DLCO: 37±17 unclear Gille 2017 France Public Prospective 45 IPF; 2002 Yes PSG, study 84% male 6 patients using Previously diagnosed cross-sectional international location Age: 68.8±8.7 oxygen therapy, OSA; unstable dyspnoea study, multi- guideline unclear FVC: 72.8±20.3 with 4 on centre (national) DLCO: 45.1±18.9 nocturnal oxygen Gu 2016b China Unclear Retrospective 23 ILD; not Yes PSG, study 48% male Only patients Not provided cross-sectional specified location Age: 57±12 with isolated study, single unclear nocturnal centre hypoxaemia were included Gundogdu Turkey Nil Prospective 38 SSc-ILD; not Yes PSG, study 8% male Not provided Other rheumatic or 2020 cross-sectional specified location Age: 51.3±11.6 chronic lung disease; study, single unclear FVC: 75.5±21.1 URTI in last 6 weeks; use centre DLCO: 55.7±17.7 of drugs affecting sleep structure (narcotic drugs, benzodiazepine) Hanci Turkey Unclear Prospective 52 ILD; not Yes PSG, study 52% male Not provided Not provided 2013b cross-sectional specified location Age: 61.4±10.4 study, single unclear FVC: 75.1±22.3 centre DLCO: 56.6±20.2 Hira 1997 India Unclear Prospective 20 ILD; clinical and Yes In-laboratory 60% male Daytime resting Not provided cross-sectional lung function PSG Age: 45.6±10.3 PaO2: study, single assessment with FVC: 42.9 76±11mmHg centre HRCT confirmation Imaizumi Japan Unclear Prospective 19 Pneumoconiosis; Yes 24-hour 100% male Daytime resting Cardiovascular and 2014 cross-sectional not specified oximetry Age: 78.6±5.2 PaO2: cerebrovascular events study, multi- FVC: 67.6±15.6 87±18mmHg within last 6 months; centre (national) 2 on home hemodialysis treatment; oxygen chronic inflammatory disease; malignancy. Kolilekas Greece Unclear Prospective 14 IPF; international Yes PSG, study 64% male Daytime resting Not provided b 2010 cross-sectional guideline without location Age: 68.3±7.8 SpO2: 95±2% study, single specifying unclear FVC: 79.9±19.8 6MWT end centre DLCO: 43.8±17.2 SpO2: 89±5% Kolilekas Greece Nil Prospective 31 IPF; 2011 Yes PSG, study 77% male Daytime resting Secondary causes of 2013 cohort study, international location Age: 68±7.9 SpO2: 96±2% lung fibrosis single centre guideline unclear

FVC: 77.6±17.1 6MWT nadir DLCO: 43.8±16 SpO2: 90±5% Lancaster USA Unclear Prospective 50 IPF; 2000 Yes In-laboratory 50% male 50% on home Not provided 2009 cross-sectional international PSG Age: 69±8.1 oxygen therapy study, single guideline FVC: 58±10.5 centre DLCO: 43±13.8 Lee 2015 Ireland Public Prospective 20 IPF; international Yes In-laboratory Age: 67.9±12.3 Patients on Active coronary artery cross-sectional guideline without PSG FVC: 82.2±15.3 long-term disease; unstable study, single specifying DLCO: 51.1±15 oxygen therapy comorbid conditions centre were excluded Margarito- UK Unclear Retrospective 437 ILD; not Yes Overnight Not provided Not provided Not provided poulos cohort study, specified oximetry 2019b number of study centre unclear Mavroudi Greece Nil Prospective 40 IPF; 2000 Yes PSG, study IPF IPF Life expectancy <6 2018 cross-sectional international location 58% male Daytime resting months; severe study, single guideline unclear Age: 69.8±8.1 SpO2: 94±3% neurological or centre Sarcoidosis: FVC: 75.6±16.1 6MWT end psychiatric disease; use clinical, SpO2: 87±5% of drugs affecting sleep laboratory, and Sarcoidosis Sarcoidosis architecture; surgical radiological 33% male Daytime resting intervention of chest or confirmation Age: 53.8±10.5 SpO2: 95±2% upper airways in last 3 FVC: 89.9±21.6 6MWT end months SpO2: 91±6% Mermigkis US Unclear Retrospective 18 IPF; 2000 and Yes In-laboratory 75% male Not provided Not provided 2007 cross-sectional 2002 PSG Age: 68.1±8.8 study, single international FVC: 65.7±10.4 centre guideline DLCO: 49.9±15.3 Mermigkis Greece Unclear Prospective 34 IPF; 2002 Yes In-laboratory 62% male Not provided Not provided 2010 cross-sectional international PSG Age: 65±10.6 study, multi- guideline FVC: 72.5±18.1 centre (national) DLCO: 53.6±20.9 Mermigkis Greece Unclear Prospective 23 IPF; 2011 Yes PSG, study 78% male Daytime resting Not provided 2013 cohort study, international location Age: 68.2±11.5 SpO2: 93±4% single centre guideline unclear FVC: 76.2±20.5 DLCO: 59.5±24.1 Midgren Sweden Public Retrospective 14 ILD; not Yes In-laboratory Age: 59±8.4 Daytime resting Not provided 1990 cross-sectional specified PSG FVC: 67±17 PaO2: study, single 65±13mmHg centre Myall UK Unclear Prospective 29 ILD; multi- Yes Home PSG Not provided Not provided Not provided 2019b cross-sectional disciplinary study, number of diagnosis study centre unclear Nicoletta Italy Unclear Cross-sectional 75 ILD; not Yes Overnight 81% male Not provided Not provided 2017b study of specified cardio- Age: 68±8.7.3 unknown design,

number of study respiratory centre unclear polygraphy Okcay USA Unclear Cross-sectional 268 IPF; international Yes Overnight 62% male None using Not provided 2010b study of guideline without oximetry Age: 68±13 nocturnal unknown design, specifying supplemental number of study oxygen centre unclear Park 2011 Korea Unclear Prospective 19 IPF: 2002 Yes 24-hour 47% male None had Not provided cross-sectional international oximetry Age: 58±11 resting study, single criteria FVC: 55.6±16.4 hypoxaemia centre CTD-ILD: DLCO: 40.8±12.5 rheumatological guidelines Pereira Portugal Nil Prospective 49 ILD; not Yes Home PSG 53% male Patients using High risk for sleep 2019 cross-sectional specified Age: 67.2±12.2 long-term disorders (BMI ≥30, study, single FVC: 86.1±18.9 oxygen therapy cranium-facial deformities centre were excluded and benzodiazepine use); Daytime resting neuromuscular diseases; PaO2: severe psychiatric 80±11mmHg disorders; non-ILD chronic lung diseases; previously diagnosed OSA Perez- Canada Public Prospective 11 ILD; clinical and Yes In-laboratory 45% male Daytime resting Not provided Padilla and cross-sectional lung function PSG Age: 51.8±16 PaO2: 1985 individual study, single DLCO: 51.6±14 69±4mmHg donation centre Pihtili Turkey Unclear Prospective 50 IPF; 2002 Yes In-laboratory 36% male Daytime resting Treatment changes in the 2013 cross-sectional international PSG Age: 53.9±12.2 PaO2: last 3 months; severe study, single guideline FVC: 85.3±23.1 77±12mmHg psychiatric or centre SSc-ILD; DLCO: 71.6±25.4 neurological disease; rheumatoglocial LRTI in the last month; guideline non-ILD chronic lung Sarcoidosis: disease; use of drugs histological affecting sleep confirmation architecture; BMI ≥30 Pillai 2012 USA Public Retrospective 54 IPF; 2000 Yes In-laboratory 65% male 17 patients Connective tissue cross-sectional international PSG Age: 69.2±7 using disease study, single guideline FVC: 64±15.9 continuous centre Other ILD; DLCO: 43.4±16.5 oxygen therapy; histological or 4 patients using HRCT exertional or confirmation nocturnal oxygen therapy Pitsiou Greece Unclear Prospective 33 IPF; 2011 Yes PSG, study 79% male Not provided Not provided 2013 cross-sectional international location study, number of guideline unclear

study centre unclear Reid 2015 Scotland Unclear Cohort study of 27 IPF; 2000 Yes Home PSG 100% male Daytime resting Not provided unknown design, international SpO2: 91±6% number of study guideline (male); 90±6% centre unclear (female) Rosenthal USA Unclear Retrospective 14 ILD; not Yes PSG, study 57% male Not provided Not provided 2018b cross-sectional specified location Age: 67.4±7.2 study, number of unclear DLCO: 39.9±0.1 study centre unclear Sarac Turkey Unclear Prospective 47 ILD; not Yes PSG, study 55% male Not provide Not provided 2016b cross-sectional specified location study, number of unclear study centre unclear Sarac Turkey Public Prospective 79 ILD; clinical Yes PSG, study IPF Not provided Upper airway pathology 2019 cross-sectional assessment with location 69% male predisposing for OSA; a study, single radiological unclear Age: 62±9.4 clinical instability; ICU centre confirmation Median FVC: 71 admission or treatment Median DLCO: 60 change in the last 3 Sarcoidosis months; severe 43% male neurological or Age: 48.7±8.4 psychiatric disease; non- Median FVC: 78 ILD chronic lung disease Median DLCO: 72 use of drugs affecting sleep architecture Singh India Unclear Prospective 48 IPF; 2002 Yes Overnight Not provided Not provided Not provided 2011b cross-sectional international oximetry study, number of guideline study centre Other ILD; unclear clinical assessment with radiological confirmation Tatsumi Japan Public Prospective 14 ILD; clinical and Yes In-laboratory Age: 63±10 Daytime resting Not provided 1989 cross-sectional lung function PSG DLCO: 62±16 PaO2: study, number of assessment with 73±9mmHg study centre radiological and unclear histological confirmation Trakada Greece Unclear Prospective 38 ILD; histological Yes In-laboratory 55% male Not provided Not provided 2003 cross-sectional confirmation PSG study, single centre Troy 2019 Australia Public Prospective 92 ILD; multi- Yes In-laboratory 55% male Daytime resting PSG conducted using cohort study, disciplinary PSG Age: 66.1±10.7 PaO2: CPAP therapy single centre diagnosis based 73±11mmHg

on international FVC: 76.6±17.1 guidelines DLCO: 54.3±15.7 Tudorache Romania Nil Prospective 23 IPF; 2013 Yes Overnight 57% male Daytime resting Alcohol consumption; 2019 cross-sectional international cardio- Age: 67.6±8.7 SpO2: 92±4% neurological or study, single guideline respiratory FVC: 71.3±16.4 13 using long- psychiatric diseases; centre polygraphy DLCO: 44.1±13.9 term oxygen serious or unstable therapy associated chronic diseases Vazquez Mexico Public Prospective 19 ILD; clinical and Yes In-laboratory 32% male Daytime resting Not provided 2011 cross-sectional lung function PSG Age: 50±13 PaO2: study, single assessment with FVC: 58±17 51±7mmHg centre radiological ± DLCO: 47±13 histological confirmation Vennelle UK Unclear Cross-sectional 38 IPF; 2000 Yes Overnight 74% male Patients using Previous diagnosis of IPF 2013b study of international cardio- Age: 68±11 oxygen therapy unknown design, guideline respiratory (male); 74±7 were excluded number of study polygraphy (female) centre unclear Zahiruddin USA Unclear Retrospective 24 IPF; ICD.9 codes Yes PSG, study 87% male Not provided Not provided 2017b cross-sectional location Age: 66±9 study, single unclear FVC: 56±21 centre DLCO: 41±14 Zhang China Public Prospective 77 ILD; not Yes In-laboratory 73% male Patients using Disease exacerbation in 2019 cross-sectional specified PSG Age: 65.4±8.4 nocturnal the last 2 months study, single FVC: 76.1±20.2 oxygen therapy centre DLCO: 53.3±19.4 were excluded

Abbreviations: 6MWT, 6-minute walk test; BMI, body mass index; CPAP, continuous positive airway pressure; CXR, chest X-ray; DLCO, diffusing capacity for carbon monoxide; FVC, forced vital capacity; HRCT, high-resolution computed tomography; ICD.9, International Classification of Diseases, Ninth Revision; ICU, intensive care unit; ILD, interstitial lung disease; IPF, idiopathic pulmonary fibrosis; LRTI, lower respiratory tract infection; OSA, obstructive sleep

apnoea; PaO2, partial pressure of oxygen; PSG, polysomnography; RCT, randomised controlled trial; SpO2: oxyhaemoglobin saturation; SSc-ILD, scleroderma- associated interstitial lung disease; UK, United Kingdom; URTI, upper respiratory tract infection; USA, United States of America

a Age, FVC, and DLCO are presented in mean ± standard deviation, unless otherwise specified b Abstract only c Data was only extracted from the abstract, given the manuscript in foreign language

Table S5. Summary of the risk of bias assessment for non-intervention studies Risk of bias assessment itemsa Risk of bias References I1 I2 I3 I4 I5 I6 I7 I8 I9 I10 I11 summary Ahmed 2018 1 0 0 0 0 0 0 0 0 0 0 Low Avdeev 2011 1 1 1 1 0 1 0 1 0 0 0 High Aydogdu 2006 1 1 1 1 0 1 0 0 0 0 0 High Bingol 2015 1 0 0 0 0 0 0 0 0 0 0 Low Bosi 2017 1 0 0 0 0 0 0 0 0 0 0 Low Bye 1984 1 0 1 1 0 1 0 0 1 0 0 High Canora 2019 0 0 0 0 0 0 0 0 0 0 0 Low Cardoso 2018 1 0 1 0 0 1 0 0 0 0 0 Moderate Chudiwal 2012 1 1 1 1 1 1 0 0 1 0 0 High Clark 2001 1 0 0 1 0 1 0 0 0 0 0 Moderate Corte 2012 1 1 1 0 1 1 0 0 0 0 0 Moderate Desideri 2013 1 1 1 1 0 1 1 1 0 1 1 High Ernst 2015 1 1 1 1 0 0 0 0 0 0 0 High Frija-Masson 2019 1 1 1 1 1 1 0 0 0 0 0 High Gille 2017 0 0 1 1 0 0 0 0 0 0 0 Moderate Gu 2016 1 1 1 1 1 1 0 0 0 0 0 High Gundogudu 2020 1 0 0 1 0 1 0 0 0 0 0 Moderate Hanci 2013 1 1 1 1 0 1 1 0 1 0 1 High Hira 1997 1 1 1 1 0 0 0 0 0 0 0 Moderate Imaizumi 2014 1 1 0 1 0 1 0 0 0 0 0 High Kolilekas 2010 1 0 1 1 0 0 0 0 0 0 0 Moderate Kolilekas 2013 0 0 0 1 0 0 0 0 0 0 0 Low Lancaster 2009 1 0 1 1 0 0 0 0 0 0 0 Moderate Lee 2015 1 0 1 1 0 0 0 0 0 0 0 Moderate Margaritopoulos 2019 1 1 1 1 1 1 0 0 0 0 0 High Mavourdi 2018 1 0 1 0 0 0 0 0 0 0 1 Moderate Mermigkis 2007 0 0 1 1 1 0 0 0 0 0 0 Moderate Mermigkis 2010 1 0 0 0 0 0 0 0 0 0 0 Low Midgren 1990 1 1 1 1 1 1 0 0 0 0 0 High Myall 2019 1 0 1 1 0 0 0 0 0 0 1 Moderate Nicoletta 2017 1 1 1 1 1 1 0 0 0 0 1 High Okcay 2010 0 1 0 1 1 0 0 0 0 0 0 Moderate Park 2011 1 1 1 1 0 0 0 0 0 0 0 Moderate Pereira 2019 1 0 1 0 0 1 0 0 0 0 0 Moderate Perez-Padilla 1985 1 1 1 1 0 1 0 0 0 0 0 Moderate Pihtili 2013 1 0 0 0 0 0 0 0 0 0 0 Low Pillai 2012 1 1 1 1 1 0 0 0 0 0 0 Moderate Pitsiou 2013 1 1 0 1 0 0 0 0 0 0 0 Moderate Reid 2015 1 1 1 1 1 0 0 0 0 0 0 Moderate Rosenthal 2018 1 1 0 0 1 1 0 0 0 0 0 High Sarac 2016 1 1 1 1 0 1 0 0 0 0 0 High Sarac 2019 1 0 1 1 0 0 0 0 0 0 0 Low Singh 2011 1 1 0 0 0 0 0 0 0 0 0 Low Tatsumi 1989 1 1 1 1 0 1 0 0 0 0 0 Moderate Trakada 2003 1 1 1 1 0 1 0 0 0 0 0 High Troy 2019 0 1 1 1 0 0 0 0 0 0 1 Low Tudorache 2019 1 1 1 1 0 0 0 0 0 0 0 Moderate Vennelle 2013 0 1 1 1 1 0 0 0 0 0 0 High Zahuruddin 2017 1 0 0 0 1 1 0 0 0 0 0 Moderate Zhang 2019 1 0 1 1 0 1 0 0 0 0 0 Moderate

a Risk of bias assessment items as described in Online Supplement Table S2. In summary: Items 1 and 2, representativeness of population; Items 3 and 4, participant selection; Items 5 and 9, data collection; Items 6 and 7, case definition; Item 8, study instrument validity; Items 10 and 11, limitations. (Scores: 1 = high risk, 0 = low risk)

Table S6. Summary of the risk of bias assessment for intervention studies

a) Non-randomised study

Risk of bias assessment items Deviation Confounding Selection Intervention from Missing Measurements Reported References Bias bias classification intervention data of outcome results Overall Mermigkis 2013 Moderate Low Low Moderate Low Moderate Low Moderate Vazquez 2001 Low Moderate Low Low Low Moderate Low Moderate

b) Randomised study

Risk of bias assessment items Carryover Deviation from Missing Measurements Reported References Randomisation effects intervention data of outcome results Overall Al Aghbari 2016 Moderate Moderate Low Low Low Moderate Moderate

Figure S1. Sensitivity analyses of pooled proportions of clinically significant nocturnal hypoxaemia: a) risk of bias, b) types of study design, and c) study participant numbers

a) Risk of bias

b) Types of study design: Retrospective studies/unclear design vs prospective studies

c) Study participant numbers: < 20 vs ≥ 20)

Table S7. Associations for clinically significant nocturnal hypoxaemia and nocturnal oxygenation parameters

Nocturnal Variables Reference oxygenation Measurement of effects p-value parameters Demographic characteristics Age Clark 2001 TST90 >2%a OR = 1.01 0.70 Non-desaturator: mean 66 vs Ockay 2010 TST90 >10%a 0.01 Desaturator: mean 70 Clark 2001 Mean SpO2 r = -0.04 0.60 Reid 2015 ODI - NSb Troy 2019 TST90 r = 0.002 0.98 BMI Non-desaturator: mean 28 vs Ockay 2010 TST90 >10%a 0.0001 Desaturator: mean 31 Bosi 2017 Mean SpO2 r = -0.46 0.007 Non-obese: mean 81.4 vs Ahmed 2018 Nadir SpO 0.66 2 obese: mean 82.33 Number of Non-obese: mean 15.2 vs Ahmed 2018 0.15 desaturation <90% obese: mean 73 Pihtili 2013 ODI r = 0.351 0.012 Reid 2015 ODI - NSb Non-obese: mean 23 vs Ahmed 2018 TST90 0.43 obese: mean 25.89 Troy 2019 TST90 r = 0.23 0.03 Sex Clark 2001 TST90 >2%a OR = 1.24 0.70 Non-desaturator: 41% male Ockay 2010 TST90 >10%a 0.02 vs Desaturator: 59% male Clark 2001 Mean SpO2 r = -0.37 0.80

ILD subtypes and disease severity ILD subtypes IPF: mean 82.4, sarcoidosis: Mavroudi 2018 Nadir SpO 0.006 2 mean 85.3 IPF: mean 83.88, b Pihtili 2013 Nadir SpO2 scleroderma: mean 85.77, NS sarcoidosis: mean 86.73 Margaritopoulos NSIP (higher ODI) vs IPF 0.03 ODI 2019 CTD-ILD (higher ODI) vs IPF 0.02 IPF: mean 13.4, scleroderma: Pihtili 2013 ODI mean 9.6, sarcoidosis: mean NSb 19.69 IPF: mean 8.98, scleroderma: Pihtili 2013 TST90 mean 4.13, sarcoidosis mean NSb 4.73 b GAP index Bosi 2017 Mean SpO2 - NS Bosi 2017 ODI - NSb Bosi 2017 TST90 - NSb CPI Limited fibrosis patients: Corte 2012 Nadir SpO2 0.93c r = -0.01 Numbers of Limited fibrosis patients: Corte 2012 0.47c desaturation <90% r = 0.10 Limited fibrosis patients: Corte 2012 ODI 0.29c r = 0.14 Troy 2019 TST90 r = 0.23 0.03

Clinical, radiological, Aydogdu 2006 Sleep SpO Negative correlation Sb,d physiological 2 scoring system ILD severity All: r = -0.502 <0.005c index IPF: - NSb,c Pihtili 2013 Mean SpO 2 Scleroderma: r = -0.705 0.001c Sarcoidosis: - NSb,c All: r = 0.370 0.008c IPF: - NSb,c Pihtili 2013 TST90 Scleroderma: r = 0.665 0.003c Sarcoidosis: - Sb,c

Pulmonary physiology Gas exchange Daytime resting Desaturation during Perez-Padilla 1985 - <0.01 SpO2 snoring b,d Aydogdu 2006 Mean SpO2 Positive correlation S c Lee 2015 Mean SpO2 r = 0.71 <0.001 b,d Aydogdu 2006 Nadir SpO2 Positive correlation S Pearson r = 0.502 0.005 Kolilekas 2013 Nadir SpO 2 Spearman r = 0.572 <0.001 c Lee 2015 Nadir SpO2 r = 0.72 <0.001 b Midgren 1990 Nadir SpO2 Highly related S c Corte 2012 Nadir SpO2 r = 0.33 0.0002 Maximum SpO Perez-Padilla 1985 2 r = 0.86 <0.01 desaturation Maximum SpO Tatsumi 1989 2 r = 0.83 <0.01 desaturation

Maximum SpO2 Perez-Padilla 1985 desaturation during r = 0.711 <0.01 Stage 2 sleep

Maximum SpO2 Perez-Padilla 1985 desaturation during r = 0.937 <0.01 REM sleep Gille 2017 ODI r = -0.24 0.12 Daytime resting Singh 2011 TST90 >10%a - 0.0015

PaO2 Lee 2015 Mean SpO2 r = 0.23 0.29 Pitsiou 2013 Mean SpO2 r = -0.259 0.18 Lee 2015 Nadir SpO2 r = 0.29 0.21 Troy 2019 TST90 r = -0.42 <0.001 Daytime resting Corte 2012 TST90 >10%a - 0.22 hypoxiae Capillary oxygen Clark 2001 TST90 >2%a OR = 0.46 0.002

concentration Clark 2001 Mean SpO2 r = 1.94 0.001 b,c 6MWT end SpO2 Corte 2012 Nadir SpO2 - NS Pearson r = 0.55 0.002 Kolilekas 2013 Nadir SpO 2 Spearman r = 0.573 <0.001 Maximum SpO Pearson r = -0.365 0.047 Kolilekas 2013 2 desaturation Spearman r = -0.338 0.068 Troy 2019 TST90 r = -0.29 0.01 6MWT Corte 2012 TST90 >10%a - 0.10 desaturation Singh 2011 TST90 >10%a - 0.0077 Exercise Non-desaturator: mean 92 vs Ockay 2010 TST90 >10%a 0.0001 desaturation Desaturator: mean 88

SpO at peak Pearson r = 0.291 0.14 2 Kolilekas 2013 Nadir SpO exercise 2 Spearman r = 0.402 0.037

DLCO Bosi 2017 Mean SpO2 r = 0.49 0.003 Mermigkis 2010 Mean SpO2 r = 0.39 0.02 Sarac 2019 Mean SpO2 r = 0.364 0.001 c Chudiwal 2012 Nadir SpO2 r = 0.629 0.016 c Kolilekas 2010 Nadir SpO2 r = 0.58 0.04 Sarac 2019 Nadir SpO2 r = 0.474 0.001 Sarac 2019 ODI r = 0.01 0.97 Sarac 2019 TST90 r = -0.686 0.001 DLCO, % Non-desaturator: mean 52 vs Ockay 2010 TST90 >10%a 0.01 predicted Desaturator: mean 47

Gu 2016 Mean SpO2 - 0.029 Pearson r = 0.467 0.01 Kolilekas 2013 Nadir SpO 2 Spearman r = 0.489 0.007 Maximum SpO Pearson r = -0.399 0.037 Kolilekas 2013 2 desaturation Spearman r = -0.373 0.046 Troy 2019 TST90 r = -0.24 0.03

KCO Frija-Masson 2019 Nadir SpO2 - 0.02 Frija-Masson 2019 ODI - 0.0007 Reid 2015 ODI - NSb

Ventilatory restriction FVC Bosi 2017 Mean SpO2 r = 0.35 0.04 b Frija-Masson 2019 Mean SpO2 - NS b Frija-Masson 2019 Nadir SpO2 - NS Frija-Masson 2019 ODI - NSb Okcay 2010 ODI - NSb Reid 2015 ODI - NSb FVC, % predicted Clark 2001 TST90 >2%a OR = 0.99 0.60

Clark 2001 Mean SpO2 r = 0.018 0.50 Troy 2019 TST90 r = -0.18 0.09 b FRC Frija-Masson 2019 Mean SpO2 - NS b Frija-Masson 2019 Nadir SpO2 - NS Frija-Masson 2019 ODI - NSb TLC Reid 2015 ODI - NSb TLC, % predicted Pearson r = 0.294 0.12 Kolilekas 2013 Nadir SpO 2 Spearman r = 0.396 0.033 Vennelle 2013 4% desaturation/hr ρ = -0.4 0.04 Vennelle 2013 ODI ρ = -0.43 0.03 Troy 2019 TST90 r = -0.04 0.68

Functional status and ventilatory responses 6MWD Singh 2011 TST90 >10%a - 0.005 c Kolilekas 2010 Nadir SpO2 r = 0.56 0.04 Pearson r = 0.463 0.01 Kolilekas 2013 Nadir SpO 2 Spearman r = 0.494 0.005 Troy 2019 TST90 r = -0.23 0.04 c VE/VCO2 Kolilekas 2010 Nadir SpO2 r = 0.72 0.012 VO peak Pearson r = 0.286 0.15 2 Kolilekas 2013 Nadir SpO 2 Spearman r = 0.377 0.05 Ventilatory Change in SpO in Tatsumi 1989 2 r = 0.77 <0.01 responses to REM sleep CO2 Change in SpO in Tatsumi 1989 2 r = 0.77 <0.01 NREM sleep

Occlusion Change in SpO in Tatsumi 1989 2 r = 0.93 <0.01 pressure REM sleep responses to Change in SpO in Tatsumi 1989 2 r = 0.90 <0.01 CO2 NREM sleep

Radiological assessments Air trapping on No air trapping: mean 80.33 Ahmed 2018 Nadir SpO 0.21 HRCT 2 vs air trapping mean 83.55 Number of No air trapping: mean 38 vs Ahmed 2018 0.42 desaturations <90% air trapping mean 75.36 No air trapping: mean 47.13 Ahmed 2018 TST90 0.14 vs air trapping mean 7.20

CXR Pihtili 2013 Mean SpO2 Negative association 0.002 involvement Pihtili 2013 ODI Direct association 0.007 Sarac 2016 TST89 Direct association 0.03 Pihtili 2013 TST90 Direct association 0.024

Warrick score Sarac 2019 Mean SpO2 r = -0.112 0.038 Sarac 2019 Nadir SpO2 r = -0.225 0.046 Sarac 2019 ODI r = 0.264 0.019 Sarac 2019 TST90 r = 0.235 0.038

Sleep-related parameters OSA TST90 ≥30% or ≥1 period of 5 minutes Mild: n=4(19%), moderate: Canora 2019 0.01 with SpO2 ≤90 with n=1(10%), severe: n=5(45%) a nadir SpO2 ≤85 ODI >10 and/or Non-OSA: n=2(13%), OSA: Cardoso 2018 0.30f TST90 > 20%a n=10(29%) Non-OSA: mean=93.1 vs Cardoso 2018 Mean SpO NSb 2 OSA: mean=92.7 No OSA: mean=93.4, Mild OSA: mean=93.6, Gille 2017 Mean SpO 0.01 2 Moderate OSA: mean=92.7, Severe OSA: mean=91.2 Non-OSA: mean 91.3, Mermigkis 2010 Mean SpO2 mild OSA: mean 90.5, 0.68 moderate-severe OSA: 89.6 Non-OSA: mean=86.1 vs Cardoso 2018 Nadir SpO NSb 2 OSA: mean=81.5) Non-OSA: mean 87.3, Lancaster 2009 Nadir SpO2 mild: mean 81.6, 0.03 moderate-severe: mean 78.7 Non-OSA: mean 84.4, Lee 2015 Nadir SpO 0.30 2 OSA: mean 81.3 Non-OSA: mean 85.6, Mermigkis 2010 Nadir SpO2 mild OSA: mean 80.2, 0.41 moderate-severe OSA: 79.6 b,c Pillai 2012 Nadir SpO2 - S Maximum SpO Non-OSA: mean 8.9, OSA: Lee 2015 2 0.30 desaturation mean 12.8 Non-OSA: mean=3.4 vs Cardoso 2018 ODI 0.009 OSA: mean=15.2 Non-OSA: mean 5.87 vs Ernst 2015 ODI <0.001 OSA: mean = 18.78

No OSA: mean=1.7, Mild OSA: mean=5.1, Moderate Gille 2017 ODI <0.0001 OSA: mean=10.1, Severe OSA: mean=50 Non-OSA: mean 4.6, mild: Lancaster 2009 ODI mean 9.3, moderate-severe: <0.001 mean 35 Non-OSA: mean 3.3, mild Mermigkis 2010 ODI OSA: mean 8.6, moderate- <0.001 severe OSA: 27.9 Non-OSA: mean 3.3, OSA: Pihtili 2013 ODI 0.001 mean 17.4 Pihtili 2013 ODI r = 0.883 <0.005 Non-OSA: mean 0.5, OSA: Lee 2015 TST86 0.037 mean 11.7 Non-OSA: mean 1.4, OSA: Lee 2015 TST88 0.02 mean 17.2 Non-OSA: mean=8.4% vs Cardoso 2018 TST90 NSb OSA: mean=16.9% No OSA: mean=4.3, Mild OSA: mean=2.9, Moderate Gille 2017 TST90 0.03 OSA: mean=9.1, Severe OSA: mean=26.8 Non-OSA: mean 2.31, OSA: Lee 2015 TST90 0.045 mean 24.15 Non-OSA: mean 12.6, mild Mermigkis 2010 TST90 OSA: mean 23.8, moderate- 0.27 severe OSA: 35.7

AHI Zhang 2019 Mean SpO2 r = -0.21 0.77 Pearson r = -0.465 0.008 Kolilekas 2013 Nadir SpO 2 Spearman r = -0.542 0.002

Mavroudi 2018 Nadir SpO2 Sarcoidosis: r = -0.821 0.001 b Pereira 2019 Nadir SpO2 - S b Pereira 2019 Nadir SpO2 - NS Zhang 2019 Nadir SpO2 r = -0.50 <0.001 Zhang 2019 Nadir SpO2 r = 0.381 0.03 Maximum SpO Chudiwal 2012 2 r = 0.508 0.027c desaturation Maximum SpO Pearson r = 0.461 0.009 Kolilekas 2013 2 desaturation Spearman r = 0.479 0.006 Mavroudi 2018 ODI r = 0.592 0.008 Mavroudi 2018 ODI Sarcoidosis: r = 0.835 0.001 Zhang 2019 ODI r = 0.91 <0.001 Zhang 2019 ODI r = 1.038 <0.001 Pereira 2019 TST90 - Sb Pereira 2019 TST90 - NSb Zhang 2019 TST90 r = 0.21 0.08

RDI Bosi 2017 Mean SpO2 r = -0.49 0.003 ODI Bosi 2017 Mean SpO2 r = -0.61 < 0.001 Standardized β = -0.26, Bosi 2017 Mean SpO2 0.002 SE β = 0.08 Mavroudi 2018 Nadir SpO2 Sarcoidosis: r = -0.851 <0.001 TST90 Bosi 2017 Mean SpO2 r = -0.89 < 0.001 Standardized β = -0.77, Bosi 2017 Mean SpO2 0.003 SE β = 0.07

Nadir SpO2 Mavroudi 2018 Mean SpO r=0.612 0.005 during sleep 2

Cardiovascular and endothelial functions Arterial stiffness Avdeev 2011 TST88 r = -0.67* <0.05 index Change in reflection index of digital volume Avdeev 2011 Mean SpO r = -0.83* <0.05 pulse in 2 response to salbutamol Ambulatory Median nighttime Imaizumi 2014 r = -0.32 <0.01c daytime SBP SpO2 b,c Imaizumi 2014 Nadir SpO2 r = 0.09 NS Imaizumi 2014 ODI r = -0.06 NSb,c Imaizumi 2014 TST90 r = 0.14 NSb,c Ambulatory Median nighttime Imaizumi 2014 r = 0.16 NSb,c nighttime SBP SpO2 b,c Imaizumi 2014 Nadir SpO2 r = -0.03 NS Imaizumi 2014 ODI r = 0.10 NSb,c Imaizumi 2014 TST90 r = -0.14 NSb,c NT-pro-BNP Troy 2019 TST90 r = 0.54 <0.0001 TRJV Limited fibrosis patients: Corte 2012 Nadir SpO 0.04 2 r = -0.34 Numbers of Limited fibrosis patients: Corte 2012 <0.05c desaturation <90% r = 0.33 Limited fibrosis patients: Corte 2012 ODI <0.05c r = 0.33 Peak tricuspid Non-desaturator: mean 2.7 Okcay 2010 TST90 >10%a 0.0008 regurgitation vs Desaturator: mean 3.02 RVSP Non-desaturator: mean 40 vs Ockay 2010 TST90 >10%a 0.003 Desaturator: mean 45 Pearson r = -0.732 <0.001 Kolilekas 2013 Nadir SpO 2 Spearman r = -0.662 <0.001 Maximum SpO Pearson r = 0.702 <0.001 Kolilekas 2013 2 desaturation Spearman r = 0.598 <0.001 Troy 2019 TST90 r = 0.43 0.001 Echocardiogra- phic estimated Pitsiou 2013 Mean SpO2 r = -0.58 0.002 sPAP Baseline pulmonary Troy 2019 TST90 ≥10%a OR 3.45 0.01 hypertension Troy 2019 TST90 ≥10%a OR 3.69 0.03 Troy 2019 Nadir SpO2 OR 0.93 0.02 Maximum SpO Troy 2019 2 OR 1.03 0.37 desaturation Troy 2019 ODI OR 1.06 <0.01 Troy 2019 TST90 OR 1.05 <0.01 New or Troy 2019 TST90 ≥10%a OR 3.59 0.03 worsening Troy 2019 TST90 ≥10%a OR 4.29 0.04 pulmonary Troy 2019 Nadir SpO2 OR 0.91 0.03 hypertension at Maximum SpO Troy 2019 2 OR 1.14 0.016 12 months desaturation

Maximum SpO Troy 2019 2 OR 1.17 0.03 desaturation Troy 2019 ODI OR 1.02 0.38 Troy 2019 TST90 OR 1.03 0.04 RV dysfunction Non-desaturator: 28% vs Ockay 2010 TST90 >10%a 0.04 Desaturator: 72%

Other c CRP Lee 2015 Mean SpO2 r = -0.66 0.019

Abbreviations: 6MWD, 6-minute walk distance; 6MWT, 6-minute walk test; AHI, apnoea-hypopnoea

index; BMI, body mass index; CO2, carbon dioxide; CPI, composite physiologic index; CRP, c-reactive protein; CTD-ILD, connective tissue disease-associated interstitial lung disease; CXR, chest X-ray; DLCO, diffusing capacity for carbon monoxide; FRC, functional residual capacity; FVC, forced vital capacity; GAP, Gender-Age-Physiology; HRCT, high-resolution computed tomography; IPF, idiopathic pulmonary fibrosis; ILD, interstitial lung disease; KCO, carbon monoxide transfer coefficient; NREM, non-rapid eye movement; NS, non-significant; NSIP, non-specific interstitial pneumonia; NT-pro-BNP, N-terminal pro b-type natriuretic peptide; ODI, oxygen desaturation index; OR, odd ratio; OSA,

obstructive sleep apnoea; PaO2, partial pressure of oxygen; RDI, respiratory disturbance index; REM, rapid eye movement; RV, right ventricular; RVSP, right ventricular systolic pressure; S, significant;

SBP, systolic blood pressure; SE, standard error; sPAP, systolic pulmonary artery pressure; SpO2, oxyhaemoglobin saturation; TLC, total lung capacity; TRJV, tricuspid regurgitant jet velocity; TST86,

percentages of total sleep time with SpO2 < 86%; TST88, percentages of total sleep time with SpO2 <

88%; TST89, percentages of total sleep time with SpO2 < 89%; TST90, percentages of total sleep time

with SpO2 < 90%; VE/VCO2, minute ventilation-carbon dioxide production relationship; VO2 peak, peak oxygen consumption

Multivariable analyses are shaded in grey. a Thresholds used for clinically significant nocturnal hypoxaemia b P-values not provided c Uncertainty about being adjusted or unadjusted analyses d Data was only extracted from the abstract, given the manuscript in foreign language e Defined as daytime resting SpO2 <92% f P-value was calculated from available data

Table S8. Impacts of supplemental oxygen therapy for nocturnal hypoxaemia

(a) Single-night in-laboratory polysomnographic study of supplemental oxygen versus room air41 Oxygen Room n p-value therapy air Polysomnography measurements Mean SpO2, % 19 94.8±2.9 82.3±9.1 <0.001 Nadir SpO2, % 19 84.7±6.3 68.7±12.3 <0.001 Heart rate, beats per minute 19 64±8.1 79.5±12.2 <0.001

(b) 1-month randomized double-blind crossover trial of supplemental oxygen versus medical air13, 66 Oxygen Medical n p-value therapy air Polysomnography measurements Percentages of total sleep time with SpO2 < 90%, % 12 6.2±15.5 37.9±35 <0.01 Oxygen desaturation index 12 3.3±6.9 12.4±5.9 <0.01 Apnoea-hypopnoea index 12 3.3±5.3 15.5±16.3 <0.01

Patient-reported outcomes Epworth Sleepiness Scale 12 - - NSa Pittsburgh Sleep Quality Index 12 - - NSa St. George's Respiratory Questionnaire 12 - - NSa UCSD Shortness of Breath Questionnaire 12 - - NSa

Blood biomarker measurements Serum B-type natriuretic peptide 12 - - NSa Plasma antioxidant thiol levels, micromole/L 11 173.8±50.7 131.0±28.0 0.03

Abbreviations: NS, non-significant; SpO2, oxyhaemoglobin saturation; UCSD, University of California San Diego

Results are presented as mean ± standard deviation

a Results for each treatment arm and p-values were not provided

Table S9. Impacts of nocturnal hypoxaemia on health-related quality of life and symptoms in ILD

Nocturnal Variables Reference oxygenation Measurement of effects p-value parameters Health-related quality of life KBILD Non-desaturator: median Myall 2019 TST90 ≥10%a 53.8 vs Desaturator: mean 0.15 56.7 SGRQb Non-desaturator: mean 34.6 Myall 2019 TST90 %a 0.09 ≥10 vs Desaturator: mean 50.0

Bosi 2019 Mean SpO2 r = -0.46 < 0.01 Bosi 2019 ODI r = 0.41 0.01 Bosi 2019 TST90 r = 0.43 0.01 CRQ-SR Clark 2001 TST90 >2%a OR = 0.75 0.30 Dyspnoea

Clark 2001 Mean SpO2 r = 0.38 0.50 CRQ-SR Fatigue Clark 2001 TST90 >2%a OR = 0.67 0.10

Clark 2001 Mean SpO2 r = 0.80 0.10 CRQ-SR Emotion Clark 2001 TST90 >2%a OR = 0.91 0.70

Clark 2001 Mean SpO2 r = 0.50 0.40 CRQ-SR Mastery Clark 2001 TST90 >2%a OR = 0.80 0.30

Clark 2001 Mean SpO2 r = 1.15 0.01 SF-36 Physical Clark 2001 TST90 >2%a OR = 0.87 0.03 Functioning

Clark 2001 Mean SpO2 r = 0.23 0.05 Mavroudi 2018 Nadir SpO2 Sarcoidosis: r = 0.462 0.035 SF-36 Role Clark 2001 TST90 >2%a OR = 0.71 0.20 Physical

Clark 2001 Mean SpO2 r = 0.60 0.20 SF-36 Role Clark 2001 TST90 >2%a OR = 0.77 0.30 Mental

Clark 2001 Mean SpO2 r = -0.16 0.80 SF-36 Social Clark 2001 TST90 >2%a OR=0.77 0.03 Functioning

Clark 2001 Mean SpO2 r = 0.60 0.01 SF-36 Pain Clark 2001 TST90 >2%a OR = 0.88 0.30

Clark 2001 Mean SpO2 r = 0.11 0.70 SF-36 Mental Clark 2001 TST90 >2%a OR = 0.94 0.30 Health

Clark 2001 Mean SpO2 r = 0.22 0.06 SF-36 Energy Clark 2001 TST90 >2%a OR = 0.77 0.005

Clark 2001 Mean SpO2 r = 0.41 0.004 SF-36 Health Clark 2001 TST90 >2%a OR = 0.84 0.04 Perception

Clark 2001 Mean SpO2 r = 0.15 0.30 SF-36 Mental Sarcoidosis: Component Mavroudi 2018 Mean SpO 0.05 2 r = -0.428 Summary

Symptoms Berlin Mavroudi 2018 ODI r = 0.484 0.036 Questionnaire Clark 2001 TST90 >2%a OR = 1.27 0.006

Epworth Clark 2001 Mean SpO2 r = -0.15 0.30 Sleepiness Scale Mavroudi 2018 Mean SpO2 r = -0.557 0.013 Maximum SpO Chudiwal 2012 2 r = 0.238 0.39c desaturation Expanded Number of Disability Status Ahmed 2018 r = 0.655 c desaturations <90% ≤0.05 Scale HADS Anxiety Clark 2001 TST90 >2%a OR = 1.04 0.60

Clark 2001 Mean SpO2 r = -0.05 0.60 HADS Clark 2001 TST90 >2%a OR = 1.13 0.10 Depression

Clark 2001 Mean SpO2 r = -0.19 0.30 c MRC score Kolilekas 2010 Nadir SpO2 r = 0.80 0.001 Pearson r = -0.477 0.007 Kolilekas 2013 Nadir SpO 2 Spearman r = -0.417 0.019 Maximum SpO Pearson r = 0.382 0.034 Kolilekas 2013 2 desaturation Spearman r = 0.269 0.14 MoCA MoCA<23: mean 66.1, Tudorache 2019 Nadir SpO2 0.65 MoCA≥23: mean 69.4 MoCA<23: mean 26.38, Tudorache 2019 ODI 0.19 MoCA≥23: mean 15.22

Abbreviations: CRQ-SR, self-reported Chronic Respiratory Questionnaire; HADS, Hospital Anxiety and Depression Scale; KBILD, King's Brief Interstitial Lung Disease questionnaire; MoCA, Montreal Cognitive Assessment; MRC, Medical Research Council Dyspnoea Scale; ODI, oxygen desaturation index; OR, odd ratio; SF-36, Short-form 36 questionnaire; SGRQ, St. George's Respiratory

Questionnaire; SpO2, oxyhaemoglobin saturation; TST90, percentages of total sleep time with SpO2 < 90%

a Thresholds used for clinically significant nocturnal hypoxaemia b Measured as total score for St. George's Respiratory Questionnaire c Uncertainty about being adjusted or unadjusted analyses

Khor YH, et al. Thorax 2021;0:1–9. doi: 10.1136/thoraxjnl-2020-216749