The role of physical activity, exercise, and adiposity on cognition: Exploring novel cognitive metrics, and neuroprotective molecules

Kim, Jeongwoon

Permalink

https://hdl.handle.net/2142/127464 Copy

Description

Title

The role of physical activity, exercise, and adiposity on cognition: Exploring novel cognitive metrics, and neuroprotective molecules

Author(s)

Kim, Jeongwoon

Issue Date

2024-12-02

Director of Research (if dissertation) or Advisor (if thesis)

Khan, Naiman A

Doctoral Committee Chair(s)

Khan, Naiman A

Committee Member(s)

• Burd, Nicholas A
• Boppart, Marni
• Motl, Robert W

Department of Study

Kinesiology & Community Health

Discipline

Kinesiology

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

Physical Activity, Exercise, Cognition, Adiposity, Multiple Sclerosis, Brain-Derived Neurotrophic Factor, Cathepsin B, Optical Coherence Tomography 

Abstract

OBJECTIVES The prevalence of inactivity and obesity in the general population is a growing concern as physical activity and exercise (PAE), and excess adiposity exerts significant influences on neurocognitive health and function in both healthy populations and in populations with multiple sclerosis (MS). While an ever-growing body of research supports the neurocognitive benefits of engaging in PAE, there are several gaps and limitations that persist in the current literature. One limitation is understanding the influence of these lifestyle factors on underutilized measures of neurocognitive function and health. Specifically, there is growing evidence suggesting intra-individual variability (IIV) and retinal morphology from optical coherence tomography (OCT) may serve as markers for neural efficiency and neuronal integrity, respectively. Another gap lies in understanding the molecular underpinnings of these lifestyle factors on cognition. While a substantial body of works have implicated brain-derived neurotrophic factor (BDNF) and cathepsin B (CTSB) as key molecules for these effects, the current evidence is inconsistent. Therefore, this dissertation aims to investigate the impact of PAE and adiposity on neurocognitive health and function by utilizing both traditional and underutilized measures of neurocognitive health and function (i.e., IIV and OCT). Furthermore, this dissertation will also examine the influence of these lifestyle factors on BDNF and CTSB to understand their role in neurocognitive health and function in both individuals with and without MS. Aim 1 examined the relationship between moderate-to-vigorous physical activity (MVPA) and retinal morphology in individuals with MS. Aim 2 explored the combined effects of a 10-week resistance training intervention and differing dietary protein intake on cognitive performance, neuroelectric patterns, BDNF, and CTSB to then examine the relationship between changes in BDNF and CTSB with changes in neurocognitive outcomes. Aim 3 explored the role of excess adiposity on both traditional measures of behavioral performance and neuroelectric patterns as well as IIV for both outcomes. Aim 4 examined the effects of an acute bout of high-intensity cycling on different domains of cognitive function and the influence of IIV on these exercise-induced effects. Aim 5 examined the time-dependent changes in BDNF and CTSB during an acute bout of high-intensity cycling compared to a resting control to assess the relationship between changes in BDNF and CTSB with changes in neurocognitive outcomes. Aim 6 investigated the associations between BDNF and CTSB with both traditional measures of behavioral performance and neuroelectric patterns as well as IIV for both outcomes. Aim 7 aimed to examine the differences in BDNF and CTSB between healthy controls and individuals with MS to examine their associations with structural integrity and cognitive outcomes. METHODS Aim 1 utilized secondary data in individuals with MS (n=41) and healthy controls (n=79) who wore ActiGraph wGT3X+ accelerometers to measure physical activity and completed retinal imaging through OCT. Aim 2 included adults (n=40) between the ages of 40 and 64 years of age. These individuals completed baseline cognitive testing consisting of a modified flanker task along with the N2 and P3 event-related potentials (ERPs) components to assess attentional inhibition, and a spatial reconstruction task to assess spatial memory. These individuals also provided fasted blood and muscle samples to quantify BDNF and CTSB. These individuals then completed 10-weeks of resistance training in combination with differing levels of dietary protein before completing the cognitive tasks and providing biological samples following its completion. Aim 3 utilized secondary data from participants across multiple projects (n=320) who completed dual-energy X-ray absorptiometry (DEXA) for whole body fat percentages and the modified flanker task with N2 and P3 ERPs for attentional inhibition. Aim 4 included adults (n=20) that completed a randomized cross-over trial consisting of 30-minutes of cycling and sedentary control. These participants completed a battery of cognitive tasks consisting of the modified flanker, incidental statistical learning, and spatial memory task before and after both conditions along with ERP recordings. Aim 5 consisted of adults (n=19) who completed the previous randomized cross-over trial and provided blood samples to quantify BDNF and CTSB at five timepoints during both the exercise and sedentary control conditions. Aim 6 consisted of adults (n=166) predominately characterized as either overweight or obese (79%) from previous studies who provided blood samples, adiposity measures from DEXA, and completed the modified flanker task with N2 and P3 ERPs. Aim 7 consisted of individuals with MS (n=24) and healthy controls (n=24) with available blood samples, retinal imaging, and modified flanker task outcomes with N2 and P3 ERPs. RESULTS For Aim 1, there were significant differences in TMV (t=4.82, p<0.01), RNFL (t=4.69, p<0.01), and MVPA (t=6.60, p<0.01), between individuals with MS and healthy controls. Furthermore, upon adjusting for age in subsequent regression models, MVPA was associated with greater TMV (β=0.34; p<0.01) and RNFL (β=0.22; p=0.03) that attenuated differences between the two groups. In analyses of Aim 2, while differing levels of protein intake did not influence BDNF, CTSB, or cognitive outcomes, the 10-weeks of resistance training resulted in increased circulating BDNF (µd=0.617; p=0.035) and muscle CTSB mRNA expression (µd=0.268; p=0.036). Regression models examining the associations between changes in BDNF and CTSB with changes in cognitive outcomes revealed changes in BDNF to be associated with faster reaction times (congruent: β=−0.38, p=0.026; incongruent: β=−0.38, p=0.024) and congruent N2 latency (β=−0.52, p=0.016) while changes in CTSB were associated with faster incongruent P3 latencies (β=−0.52, p=0.016). The linear regression analyses for Aim 3 revealed that, upon adjusting for age and sex, whole body fat percentage was associated with decreased accuracy (congruent: β=−0.18, p=0.016; incongruent: β=−0.23, p<0.001) , greater reaction time variability (congruent: β=0.13, p=0.032; incongruent: β=0.23, p<0.001), slower P3 latencies (congruent: β=0.21, p=0.003; incongruent: β=0.18, p=0.010), greater incongruent N2 latency variability (β=0.16, p=0.017) , and greater incongruent P3 latency variability (β=0.16, p=0.025). For Aim 4, mixed models revealed no exercise-specific effects for flanker performance or neuroelectric patterns, no effects for spatial reconstruction performance, and negative effects, characterized by reduced accuracy (F=5.47; p=0.040), slower reaction times (F=5.18; p=0.036), and smaller late parietal positivity amplitudes (F=4.26; p=0.046) for the implicit statistical learning task. However, upon adjusting for IIV, there were exercise-specific decreases in reaction time (F=24.00; p<0.001), reductions in incongruent N2 amplitudes (F=13.03; p=0.002) and slower incongruent P3 latencies (F=3.57; p = 0.065). For Aim 5, mixed models revealed time-dependent increases in BDNF (F=11.27, p<0.001) and CTSB (F=3.33, p=0.022) during exercise compared to sedentary control. Subsequent analyses utilizing area under the curve for BDNF and CTSB revealed greater increases in BDNF to be associated with faster P3 latencies during the congruent (F=3.16, p=0.099) but slower during the incongruent (F=7.41, p=0.014) trials. Upon adjusting for age, sex, and whole-body fat percentage, linear regression models in Aim 6 revealed CTSB to be associated with more negative N2 amplitudes (congruent: ꞵ=-0.17, p=0.036; incongruent: ꞵ=-0.14, p=0.083) while BDNF was associated with lower congruent N2 latency variability (ꞵ=-0.15, p=0.045). For Aim 7, independent t-test revealed no differences in either BDNF (t=0.24, p=0.811) or CTSB (t=0.81, p=0.422) between individuals with MS and healthy controls. Subsequent linear regression analyses revealed BDNF to be associated with poorer congruent accuracy (ꞵ=-0.29, p=0.035), slower reaction times (congruent: ꞵ=0.42, p=0.001; incongruent: (ꞵ=0.37, p=0.003), and greater congruent reaction time variability (ꞵ=0.30, p=0.040). However, neither BDNF nor CTSB were associated with retinal morphology outcomes, nor the N2 or P3 ERP components. CONCLUSION These works provide preliminary evidence for the utility of underutilized neurocognitive outcomes that underscore the influences of PAE and adiposity on neurocognitive health and function. Additionally, these works reveal insights into how different PAE interventions, adiposity, and MS influence BDNF and CTSB concentrations and their associations with both established and novel cognitive metrics. Specifically, the findings in Aim 1 suggest that greater MVPA was linked to greater RNFL thickness and TMV. Importantly, these associations attenuated retinal morphology differences between individuals with MS and healthy control. Accordingly, the use of MVPA may be a viable therapeutic option in maintaining retinal neuronal and axonal integrity in MS. From the findings in Aim 2, the association between increases in BDNF and CTSB with improvements in cognitive outcomes following 10-weeks of resistance training may suggest increases in BDNF and CTSB are necessary to observe cognitive improvements following lifestyle interventions. The associations between whole-body adiposity and increased variability across both behavioral and ERPs from Aim 3 not only confirm the detrimental effects of excess adiposity on neurocognitive function but also suggest that these effects extend to markers of neuronal efficiency. In Aim 4, the differing effects of an acute bout of high intensity cycling on different domains of cognitive function suggests that the cognitive benefits of acute bouts of exercise are not universal. Furthermore, as the adjustment of IIV revealed exercise-specific improvements, these findings further highlight the importance of IIV and suggest it may represent a specific cognitive adaptation to exercise. The findings in Aim 5 highlight a key gap in previous works that examined the effects of an acute bout of exercise on circulating molecules through a pre-post design and how these changes might relate to cognitive changes. The results from Aim 6 further strengthen the evidence for BDNF and CTSB in regulating neurocognitive function and suggests these molecules may explain measures of neurocognitive functions beyond what can be explained through aging and adiposity. Finally, the results of Aim 7 suggest that circulating concentrations of BDNF and CTSB may not differ between individuals with MS and healthy controls and that BDNF may negatively impact cognitive outcomes. However, the influence of MS on BDNF concentrations must be explored to understand their source and role in neurocognitive health and function. While future interventions and longitudinal works are necessary to better contextualize the findings from these works, the preliminary evidence presented from this dissertation allow researchers and clinicians to be better informed when designing and implementing interventions to provide neurocognitive benefits in both healthy populations and the MS population.

Graduation Semester

2024-12

Type of Resource

Thesis

Handle URL

https://hdl.handle.net/2142/127464

Copyright and License Information

Copyright 2024 Jeongwoon Kim

Owning Collections