Digitala Vetenskapliga Arkivet

Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Prediction of treatment efficacy in patients with obstructive sleep apnoea and type 2 diabetes: a retrospective cohort study
Linköping University, Department of Computer and Information Science, The Division of Statistics and Machine Learning.
2025 (English)Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesis
Abstract [en]

Type 2 diabetes is a common illness in which a person either does not produce enough insulin, or the body is unable to absorb insulin into the blood, and which often leads to serious complications such as several types of cardiovascular disease and disorders in heartfunction, stroke, and many others. It is highly comorbid with obstructive sleep apnoea, a sleep disorder in which people experience episodes of reduced or stopped breathing during sleep, and which causes an increased risk for cardiovascular disease. Besides weight loss, continuous positive airway pressure is the first choice of treatment for patients with obstructive sleep apnoea. Little is known about the relationship between continuous positive airway pressure treatment and risk of cardiovascular disease in patients with both obstructive sleep apnoea and type 2 diabetes. This thesis aims to explore this relationship using data from several Swedish health care registers, including the National Diabetes Register and the Swedish Sleep Apnoea Register. As these registers may grow large, this thesis additionally attempts to find methods to explore this relationship in a big data setting. It does so by modelling survival time with respect to all-cause mortality, acute myocardial infarction, and stroke using the Cox proportional hazards model and a big data divide-and-conquer adaptation of the Cox proportional hazards model in two groups. The first group includes everyone with type 2 diabetes in National Diabetes Register, a total of 750,299 people. While only 18,878 of these people also appear in Swedish Sleep Apnoea Register, many people will likely have undiagnosed obstructive sleep apnoea. The second group is the intersection between people with type 2 diabetes in National Diabetes Register and Swedish Sleep Apnoea Register. In this group, it is known that everyone has both type 2 diabetes and obstructive sleep apnoea, though it includes only 14,138 people.The results from the models were inconclusive as to the effect of continuous positive airway pressure treatment. Conflicting results were found between groups and between time-independent and time-dependent analyses, some pointing towards continuous positive airway pressure being protective of all-cause mortality, and some estimating no effect or an adverse effect. For acute myocardial infarction and stroke, similar conflicting results were found. Further work is necessary to see whether an effect truly exists, perhaps usinga causal framework. The use of antithrombotic agents was consistently estimated as having an adverse effect on survival time, a finding which may be explained by underlying comorbidities that are correlated with people taking the drugs. The divide-and-conquer algorithm was insufficient in modelling complex function involving splines, often running into numerical issues.

Place, publisher, year, edition, pages
2025. , p. 88
Keywords [en]
statistics, type 2 diabetes, obstructive sleep apnoea, survival analysis
National Category
Medical Biostatistics
Identifiers
URN: urn:nbn:se:liu:diva-216063ISRN: LIU-IDA/STAT-A--25/009--SEOAI: oai:DiVA.org:liu-216063DiVA, id: diva2:1983587
Subject / course
Statistics
Supervisors
Examiners
Available from: 2025-08-11 Created: 2025-07-11 Last updated: 2025-08-11Bibliographically approved

Open Access in DiVA

fulltext(1870 kB)171 downloads
File information
File name FULLTEXT01.pdfFile size 1870 kBChecksum SHA-512
e187d56629b58ae0ce4781295f1a13dc2b71f0f5c63fb91ac43724dc251734a08ca5346887bcb692f7ec32b464013f70199d9f0abefdaba86799bc6c261e00e7
Type fulltextMimetype application/pdf

By organisation
The Division of Statistics and Machine Learning
Medical Biostatistics

Search outside of DiVA

GoogleGoogle Scholar
Total: 172 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

urn-nbn

Altmetric score

urn-nbn
Total: 419 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf