Document Type


Date of Degree

Fall 2018

Access Restrictions

Access restricted until 01/31/2020

Degree Name

PhD (Doctor of Philosophy)

Degree In


First Advisor

Voss, Michelle W.

First Committee Member

Tranel, Daniel T.

Second Committee Member

LaLumiere, Ryan T.

Third Committee Member

Denburg, Natalie

Fourth Committee Member

Vaidya, Jatin


Given the proliferating aging population, increasing moderate-to-vigorous physical activity (MVPA) and exercise is critical because it enhances overall well-being and reduces the rates of many adverse age-related health conditions. However, intervention efforts to establish sustained changes in MVPA have yielded limited success due their sole focus on conscious factors (e.g., changing goals and intentions). Thus, older adults continue to represent the highest proportion of sedentary adults, despite their knowledge of the widespread health benefits associated with PA and exercise or even their intentions to engage in such behaviors. Consequently, the benefits of PA and exercise are not being fully realized and health problems perpetuate. Developing evidence-based interventions that establish sustained changes in MVPA and exercise behaviors in older adults is a major public health priority, but this requires going beyond social-cognitive constructs. The broad goal of my dissertation is to advance scientific understanding about the neural systems associated with changes in MVPA behaviors among sedentary older adults.

Emerging evidence indicates that nonconscious processes also regulate exercise behaviors, stemming from accumulated affective responses from past exercise experiences (e.g., pleasant vs unpleasant). Grounded by current understanding about affect's role on decision-making, the present study hypothesized that physiological changes induced by single bouts of PA serve as somatic markers in the brain that guide future PA-related behaviors. Specifically, my dissertation extended previous research by testing whether acute exercise responses in affect-related brain systems predict PA behavior change following a 3-month exercise intervention. This hypothesis is supported by prior research indicating that self-reported changes in affect (i.e., pleasant/unpleasant feelings) during moderate-intensity exercise reliably predicts future MVPA behavior. The results of my dissertation advances previous findings by investigating how these affective responses to exercise are represented in the brain and how they relate to PA behavior change. I tested my central hypothesis through the following two specific aims:

In Specific Aim 1, I investigated whether the acute physiological and neural responses to exercise were related to the subjective experience of exercise in older adults. Healthy, low-active, older participants (N = 34, Age = 67.2 years, 21 females) completed an acute exercise procedure consisting of two within-subjects exercise conditions occurring on separate counter-balanced sessions. During the active condition, participants cycled at a moderate intensity (65% of maximum heart rate), and during the passive condition, their legs were moved by motorized pedals on the same machine and at the same pedal rate as in the active condition. To investigate exercise-related changes in brain function, functional MRI scans were acquired before and after the acute exercise. Additionally, salivary samples were collected throughout the experiment to provide objective biomarkers that have been linked with psychological changes. Finally, participants provided self-reported changes in affect. I found that acute exercise was associated with increases in salivary markers of sympathetic activity and decreases in salivary cortisol levels with no significant differences between conditions. Acute exercise also resulted in observable increases in positive affect with no differences between conditions. Finally, no observable acute exercise-related changes in functional connectivity occurred.

In Specific Aim 2, I identified predictors of exercise behavior change from objectively-measured biomarkers and neural systems that are acutely responsive to exercise. After completing the acute exercise sessions, participants began a 3-month supervised exercise intervention. To assess intervention-related changes in unsupervised PA behavior, participants wore a PA monitoring device for 7 days before and immediately after the intervention. Individuals who exhibited a stronger acute functional connectivity response between nucleus accumbens (NAcc) and medial prefrontal cortex (mPFC) before the intervention were more likely to increase their unsupervised levels of MVPA after the intervention. Given the role of NAcc-mPFC circuitry in affect-based decision making and self-referential processing, the present findings suggest that enhanced cognitive appraisal and awareness of affective changes are related to more sustained changes in long term behavior. This study is the first to demonstrate neurobiological evidence supporting the relationship between positive affective responses to acute exercise and long-term changes in exercise behavior. This research advocates the utility of affect-based measures in tailoring exercise interventions for sedentary older adults.


acute exercise, affect, behavior change, brain, emotions


xiii, 239 pages


Includes bibliographical references (pages 209-239).


Copyright © 2018 Timothy Benjamin Weng

Available for download on Friday, January 31, 2020

Included in

Psychology Commons