Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form

A Functional Data Analysis of Accelerometer-Based Activity

Capps, Chandler (2020) A Functional Data Analysis of Accelerometer-Based Activity. Master's Thesis, University of Pittsburgh. (Unpublished)

This is the latest version of this item.

[img]
Preview
PDF
Submitted Version

Download (625kB) | Preview

Abstract

Introduction: Depression is prevalent, often comorbid with other negative health outcomes, and can have severe direct and indirect consequences. It is critical, then, that research contributes to the identification and prediction of depression cases. Accordingly, scores of studies have shown
statistically significant associations between certain rest-activity rhythm (RAR) characteristics and current presence of depression symptoms and future depression development. However, the study of age’s moderation of the effects of the presence of depression symptoms on RARs has yet to be
thoroughly explored. This study attempts to provide insight into the interaction between age and depression status in predicting binned log activity counts and to offer a unique exploratory analysis via functional methods.
Methods: This study focuses on functional linear modelling, in particular, functional ANOVA(fANOVA). The model’s response variable is a vector of functions constructed through the interpolation and penalized smoothing of the data. These functions map time to mean log activity
count. The predictors are di- or polychotomous and include depression status, age group, gender, and ethnicity. fANOVA yields least-squares estimates of the functional effects of each covariate on the outcome. A permutation F-test determines whether the full model, including depression and age group interactions, explains significantly more variation in the data than the main-effects
model.
Results: The fANOVA yielded statistically significant results at a .05 significance level for all covariates. In particular, depression’s negative effect on mean log activity count is statistically significant and pronounced for all waking hours (approximately 6:00 am to 10:00 pm). The young adulthood effect is similar while the pronounced effects of later adulthood and older adulthood are
positive. The sole statistically significant interaction effect, depression and young adulthood, is positive and mitigates much of the effect of depression. The permutation F-test determined the full model to be statistically significant at a .05 significance level.
Conclusion: The evidence suggests that depression is statistically significantly associated with activity over typical waking hours, and the effect is moderated by age in that young adults experience a considerably lessened depression effect.
Public Health Significance: This study contributes to the understanding of the observable
manifestation of depression in adults.


Share

Citation/Export:
Social Networking:
Share |

Details

Item Type: University of Pittsburgh ETD
Status: Unpublished
Creators/Authors:
CreatorsEmailPitt UsernameORCID
Capps, Chandlerchandlerscapps@gmail.comchc2890000-0002-0868-4799
ETD Committee:
TitleMemberEmail AddressPitt UsernameORCID
Thesis AdvisorKrafty, Robertrkrafty@pitt.edu
Committee MemberKang, Chaeryoncrkang@pitt.edu
Committee MemberSmagula, Stephensmagulasf@upmc.edu
Date: 30 July 2020
Date Type: Publication
Defense Date: 12 June 2020
Approval Date: 30 July 2020
Submission Date: 12 June 2020
Access Restriction: No restriction; Release the ETD for access worldwide immediately.
Number of Pages: 42
Institution: University of Pittsburgh
Schools and Programs: Graduate School of Public Health > Biostatistics
Degree: MS - Master of Science
Thesis Type: Master's Thesis
Refereed: Yes
Uncontrolled Keywords: Rest-activity rhythm
Date Deposited: 31 Jul 2020 02:09
Last Modified: 31 Jul 2020 02:09
URI: http://d-scholarship.pitt.edu/id/eprint/39356

Available Versions of this Item

  • A Functional Data Analysis of Accelerometer-Based Activity. (deposited 31 Jul 2020 02:09) [Currently Displayed]

Metrics

Monthly Views for the past 3 years

Plum Analytics


Actions (login required)

View Item View Item