The present project aims to that examine the interplay between sleep duration circadian phase, and metabolic regulation. The impact of CPAP therapy on metabolic regulation is assessed with a RCT. Patients were enrolled form referral to sleep clinic for suspect obstructive sleep apnea (OSA). Subjects with hypopnea index (AHI) >=15/h are randomly allocated to the usual care or Auto-PAP (APAP) therapy then reevaluated after 12th week. The CPAP effect are measured with the comparison of changes in outcomes between usual care and APAP therapy.

Enrolled participants undertake pre-randomization evaluations. That included standard questionnaire, eating behavior, International Physical Activity Questionnaire (IPAQ)-short form, 3-day dietary intake, and 7-day sleep log, 24-hr ambulatory blood pressure, and 7-day Actiwatch. Within one week after completion, participants underwent repeated polysomnography (PSG) to confirm the diagnosis of OSA. The basal metabolic rate (BMR) and body composition (BC), and blood sampling are conducted in the morning next to overnight PSG and BMR.

Once the pre-randomization evaluation was completed and the AHI >=15/h was confirmed, participants are randomly allocated to the usual care or auto-CPAP treatment (AirSense™ 10 Autoset™, ResMed Inc., Australia) for 12 weeks. Participants had two follow-up visits at 4th and 12th week and were reevaluated with the same measurements as pre-randomization Enrolled participants are randomized into CPAP or usual care. The randomization is generated by the staff of the Department of Biostatistics independent of subject enrolment. A block-randomized assignment with a block size of four is used. Consecutive patients are assigned to the allocation according to the randomization schedule by the study nurse. Investigators who assess outcome are blinded to the allocation assignment. Preset-dropout criteria: Defaulting clinic visits at 4th and 12-week, re-assessment, and clinical exacerbations requiring medical attention.

Participants starting CPAP are instructed to use the device for >=6hr/night. PAP compliance is assessed as the average usage hours per night which is objectively recorded on the devices with a built-in compliance meter. Compliance data are downloaded by medical staff at every clinic visit. The good compliance is defined as usage>= 4hr/night in 70% days of total period.

Inclusion Criteria:

1. >=20 year-old
2. patients with moderate-severe OSA (AHI≥15/hr)

Exclusion Criteria:

1. skin allergy
2. wrist tattoos
3. BMI≧40 kg/m2
4. Neurological Disorders(Infection, Stroke, ALS, Myopathy, Brain tumor and Encephalitis)
5. Internal Diseases(cardiovascular disease or coronary artery disease was poorly controlled
6. chronic heart failure(including NYHA class 3 or more than NYHA class 3)
7. COPD(FEV1/FVC<70)
8. Hyperthyroidism or Hypothyroidism
9. Primary Aldosteronism
10. Chronic Kidney Disease(eGFR<30)
11. Acromegaly and Parkinson's disease)
12. Psychosis(Schizophrenia, bipolar disorder and depression)
13. Autoimmune disease(System Lupus Erythematosus, Rheumatoid arthritis, Sarcoidosis, Crohn's disease and Ulcerative Colitis)
14. Cancer(in last 5 years)
15. Have been diagnosed with obstructive sleep apnea
16. other sleep disorder(Narcolepsy, insomnia and RBD)
17. Use anti-inflammatory drugs(in last 2 weeks)
18. Illiterate
19. shift worker
20. has been Spinal Surgery and move difficult
21. professional driver has severe drowsiness