Modeling developmental trajectories with nonrandomly missing data: investigating trajectories of frailty using data from the Manitoba Follow-up Study
| dc.contributor.author | Li, Chendong | |
| dc.contributor.examiningcommittee | Tate, Robert (Community Health Sciences) | |
| dc.contributor.examiningcommittee | St. John, Philip (Internal Medicine) | |
| dc.contributor.supervisor | Jiang, Depeng | |
| dc.date.accessioned | 2023-08-25T19:28:31Z | |
| dc.date.available | 2023-08-25T19:28:31Z | |
| dc.date.issued | 2023-08-21 | |
| dc.date.submitted | 2023-08-21T13:47:18Z | en_US |
| dc.date.submitted | 2023-08-25T18:53:15Z | en_US |
| dc.degree.discipline | Community Health Sciences | en_US |
| dc.degree.level | Master of Science (M.Sc.) | |
| dc.description.abstract | Frailty is an age-related syndrome, marked by declines in various organ systems. Its baseline and progress are linked with increased mortality. Notably, frailty trajectories exhibit significant individual variability. Understanding these trajectories and their associated factors is vital. The study assesses conventional and extended Group-based trajectory analysis (GBTA) in mapping frailty trajectories, especially considering different missing data mechanisms. Objectives: Compare conventional and extended GBTA under various conditions. Explore frailty trajectories in older men using the Manitoba Follow-up Study (MFUS) data. Methods: Simulation studies were conducted to contrast the GBTA models. Data were crafted from distinct trajectory scenarios and missing data mechanisms. Metrics like absolute error and mean squared error gauged the models' accuracy. MFUS data, a Canadian longitudinal study on ageing, was employed for real-world frailty trajectory examination. A frailty index using daily living activities and comorbidities was established, measuring health variation. Results: Simulations showed the extended GBTA was superior when latent classes weren’t initially distinct, under varied missing data scenarios. With the MFUS data, the extended GBTA identified four frailty trajectories. Participants' age correlated with their frailty trajectory, but marital status and living arrangements didn’t. Conclusion: The extended GBTA outperforms the conventional method, especially with latent classes not distinctly separated initially. By leveraging both GBTAs on MFUS data, we gain insight into frailty's growth heterogeneity. This aids in devising impactful future prevention initiatives by understanding the nuances of frailty trajectories. | |
| dc.description.note | October 2023 | |
| dc.identifier.uri | http://hdl.handle.net/1993/37493 | |
| dc.language.iso | eng | |
| dc.rights | open access | en_US |
| dc.subject | Group-Based Trajectory Analysis | |
| dc.subject | Missing Data | |
| dc.subject | Frailty Index | |
| dc.subject | Frailty Trajectory | |
| dc.subject | Manitoba Follow-up Study | |
| dc.title | Modeling developmental trajectories with nonrandomly missing data: investigating trajectories of frailty using data from the Manitoba Follow-up Study | |
| dc.type | master thesis | en_US |
| local.subject.manitoba | yes |