Cognitive and Structural Brain Correlates of Physical Activity and Fitness in Aging

MISSING IMAGE

Material Information

Title:
Cognitive and Structural Brain Correlates of Physical Activity and Fitness in Aging A Cross-Sectional Voxel-Based Morphometry Study
Physical Description:
1 online resource (58 p.)
Language:
english
Creator:
Zlatar, Zvinka Z
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Psychology, Clinical and Health Psychology
Committee Chair:
Crosson, Bruce A
Committee Members:
Marsiske, Michael
Anton, Stephen Douglas
Manini, Todd M

Subjects

Subjects / Keywords:
aging -- brain -- cognition -- fitness -- imaging
Clinical and Health Psychology -- Dissertations, Academic -- UF
Genre:
Psychology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract:
The current study investigated whether fitness level or self-reported physical activity is a better predictor of gray matter volume in older adults (N=29). Given the scarcity of brain imaging studies reporting a direct relationship between cognitive and morphometric brain measures, the direct relationship between executive function and gray matter volume was also assessed using voxel-based morphometry (VBM). We further analyzed data from an independent physical activity intervention study (LIFE-P), conducted for 1-year in sedentary older adults, in an attempt to support our cross-sectional findings (N=20). It was found that fitness level, and not self-reported physical activity, displayed a protective role on gray matter volume as age increased, in regions typically affected by the aging process, such as the cerebellum bilaterally, fusiform/inferior lateral occipital cortex/lingual gyrus, and left hippocampus and parahippocampal gyrus. Executive functions did not directly predict gray matter volume in any area of the brain, nor did gray matter volume predict executive function performance above and beyond the contribution of age. Findings from the LIFE-P did not support our cross-sectional results and rather suggested that there was not a long-term protective role of adherence to a physical activity regimen on the gray matter. Hence fitness level moderated the relationship between age and gray matter volume in areas previously reported to decline as a function of age, while self-reported physical activity levels failed to detect this finding. Future exercise intervention studies should rely on objective measures of fitness level (i.e., VO2max) rather than on self-reported measures of physical activity.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by Zvinka Z Zlatar.
Thesis:
Thesis (Ph.D.)--University of Florida, 2012.
Local:
Adviser: Crosson, Bruce A.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2014-08-31

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2012
System ID:
UFE0043375:00001


This item is only available as the following downloads:


Full Text

PAGE 1

1 COGNITIVE AND STRUCTURAL BRAIN CORRELATES OF PHYSICAL ACTIVITY AND FITNESS IN AGING: A CROSS SECTIONAL VOXEL BASED MORPHOMETRY STUDY By ZVINKA ZOE ZLATAR A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2012

PAGE 2

2 201 2 Zvinka Zoe Zlatar

PAGE 3

3 To my parents and wonderful family

PAGE 4

4 ACKNOWLEDGMENTS I thank my parents for their unconditional support. I would also like to thank Joseph Dzierzewski for his emotional and intellectual support and contributions to this dissertation study. I thank the Claude D. Pepper Older American Independence Center (OAIC ) and National Institute on Aging for providing me with the necessary funding to carry out this study (Supplement P30 AG 028740 & Institutional Training Grant T32 AG 020499). I also thank Drs. Bruce Crosson, Ph.D., Michael Marsiske, Ph.D., and Todd Manini, Ph.D. for their intellectual contributions. This dissertation could not have been carried out without the help of Keith McGregor, Ph.D., Andrew Bauer, and Stephanie Phan.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 ABSTRACT ................................ ................................ ................................ ..................... 9 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 11 Age Related Cognitive Decline ................................ ................................ ............... 11 Exercise and Cognition in Late Life ................................ ................................ ........ 12 Exercise and Brain Function in Late Life ................................ ................................ 13 Exercise and Brain Structure in Late Life ................................ ................................ 14 The Present Study ................................ ................................ ................................ .. 17 2 AIMS & HYPOTHESES ................................ ................................ .......................... 18 Aim 1 ................................ ................................ ................................ ....................... 18 Hypothesis 1 ................................ ................................ ................................ ........... 18 Aim 2 ................................ ................................ ................................ ....................... 18 Hypothesis 2 ................................ ................................ ................................ ........... 18 Aim 3 ................................ ................................ ................................ ....................... 19 Hypothesis 3 ................................ ................................ ................................ ........... 19 Aim 4 ................................ ................................ ................................ ....................... 19 Hypothesis 4 ................................ ................................ ................................ ........... 19 3 CROSS SECTIONAL STUDY DATA ................................ ................................ ...... 20 Meth ods ................................ ................................ ................................ .................. 20 Participants ................................ ................................ ................................ ....... 20 Exclusion Criteria ................................ ................................ ............................. 21 Aerobic Fitness and Physical Activity Assessment ................................ ........... 21 Neuropsychological Assessment ................................ ................................ ...... 22 St ructural Imaging ................................ ................................ ............................ 22 Procedure ................................ ................................ ................................ ............... 23 Data Analyses ................................ ................................ ................................ ......... 24 Structural Imaging ................................ ................................ ............................ 24 Image preprocessing ................................ ................................ ................. 24 Image analyses: regression models ................................ ........................... 25 Neuropsychological Data ................................ ................................ .................. 26 Results ................................ ................................ ................................ .................... 27

PAGE 6

6 Fitness Regression in FS L ................................ ................................ ............... 27 Fitness Regressions in PASW ................................ ................................ .......... 27 Physical Activity Regression in FSL ................................ ................................ 29 Direct Comparison between Fitness age & Physical Activity in FSL ................... 29 Neuropsychological Data ................................ ................................ .................. 30 Discussion ................................ ................................ ................................ .............. 32 4 LIFE PILOT STUDY DATA ................................ ................................ ..................... 41 Methods ................................ ................................ ................................ .................. 41 The LIFE P Study ................................ ................................ ............................. 41 Physical activit y intervention (PA) ................................ .............................. 41 Successful aging intervention (SA) ................................ ............................ 42 LIFE P inclusion criteria ................................ ................................ ............. 42 The LIFE P Follow up Brain Imaging Study ................................ ..................... 42 The Current Brain Imaging Study ................................ ................................ ..... 44 Participants ................................ ................................ ................................ 44 Data Analyses ................................ ................................ ................................ ......... 45 Behavioral Analysis ................................ ................................ .......................... 45 Structural I mages ................................ ................................ ............................. 45 Results ................................ ................................ ................................ .................... 46 Group Comparisons: Demographics and Performance ................................ .... 46 Group Comparisons: Gray Matter Volume ................................ ....................... 46 Discussion ................................ ................................ ................................ .............. 46 5 GENERAL DISCUSSI ON ................................ ................................ ....................... 49 APPENDIX ................................ ................................ ................................ .................... 51 REFERENCE LIST ................................ ................................ ................................ ........ 53 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 58

PAGE 7

7 LIST OF TABLES Table page 3 1 Cross sectional study participant characteristics ................................ ................ 37 3 2 Cross sectional study participant characteristics by median split ....................... 37 4 1 LIFE P participant characteristics ................................ ................................ ....... 48

PAGE 8

8 LIST OF FIGURES Figure page 3 1 Age X fitness interaction: Significant clusters ................................ ..................... 38 3 2 Age X fitness interaction: Overlap between two regression models ................... 39 3 3 Age X fitness interaction: Scatter plots ................................ ............................... 40 6 1 Median effect sizes per region of interest (cross sectional study) ...................... 51 6 2 Median effect sizes per region of interest (LIFE P) ................................ ............. 52

PAGE 9

9 A bstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy COGNITIVE AND STRUCTURAL BRAIN CORRELATES OF PHYSICAL ACTIVITY AND FITNESS IN AGING: A CROSS SECTIONAL VOXEL BASED MORPHOMETRY STUDY By Zvinka Zoe Zlatar May 2012 Chair: Bruce Crosson Major: Clinical Psychology The current study investigated whether fitness level or self reported physical activity is a better predictor of gray matter volume in older adults (N=29). Given the scarcity of brain imaging studies reporting a direct relationship between cognitive and mo rphometric brain measures, the direct relationship between executive function and gray matter volume was also assessed using voxel based morphometry (VBM). We further analyzed data from an independent physical activity intervention study (LIFE P), conducte d for 1 year in sedentary older adults, in an attempt to support our cross sectional findings (N=20). It was found that fitness level, and not self reported physical activity, displayed a protective role on gray matter volume as age increased in regions t ypically affected by the aging process, such as the cerebellum bilaterally fusiform/inferior lateral occipital cortex/lingual gyrus, and left hippocampus and parahippocampal gyrus. Executive functions did not directly predict gray matter volume in any are a of the brain, nor did gray matter volume predict executive function performance above and beyond the contribution of age. Findings from the LIFE P did not support our cross sectional results and rather suggested that there was not a long

PAGE 10

10 term protective role of adherence to a physical activity regimen on the gray matter. Hence fitness level moderated the relationship between age and gray matter volume in areas previously reported to decline as a function of age, while self reported physical activity level s failed to detect this finding. Future exercise intervention studies should rely on objective measures of fitness level (i.e., VO2 max ) rather than on self reported measures of physical activity.

PAGE 11

11 CHAPTER 1 INTRODUCTION Biological a ging is characterized by declines in central nervous system integrity function, and cognitive performance. The extant literature has reported an association between ex ercise, increased central nervous system integrity, and im proved cognition in late life The overarching goal of this study is to examine the influence of self reported physical activity level, fitness level, and cognitive performance on gray matter volume in older adults. Age Related Cognitive Decline As age increases, individuals display slowed reaction times on tasks measuring memory sea rch, lexical decision, mental rotation, speech discrimination, visual search, and digit symbol substitution among others (Hofer & Alwin, 2008) Above and beyond decrements in speeded tasks, older adults display reductions in memory and reasoning beginning early in adulthood (Salthouse, 2001) Further, word retrieval (Obler & Albert, 1980) and verbal fluency decline in old age (Mayr & Kliegl, 2000) disrupting effective communications (Kempler & Zelinski, 1994) Nevertheless, age related decrements in execut ive functions have been highlighted by many researchers as one of the core features of cognitive aging (Dempster, 1992; West, 1996) Executive functions can be defined as the mental processes controlling coordination, monitoring, inhibitory control, and ta sk switching (Hofer & Alwin, 2008) which are differentially affected by biological aging (West, 1996) Age related cognitive decline has been identified as a risk factor for nursing home entry and the development of degenerative disease, such as Alzheimer (Statistics, 2006; Wilson, 2002) Therefore, developing interventions to mitigate age

PAGE 12

12 important public heath goal. Exercise and physical activity level have been associated with improved cognitive function and positive emotional and physical health outcomes in older adults (Colcombe et al., 2003; McAuley, Shaffer, & Rudolph, 1995) Thus, exercise and increased physical activity levels may abate the deleterious eff ects of cognitive aging and merits further investigation. Exercise and Cognition in Late Life The relationship between exercise and cognition has been studied for decades dating back to studies by Spirduso and colleagues who compared reaction time measures in younger and older adult sportsman to that of sedentary older adults. They found that simple and choice reaction time was faster in older adults who engaged in regular exercise when compared to their se dentary counterparts (Spirduso, 1975; Spirduso & Clifford, 1978) Since then, several research groups have continued to investigate the relationship between exercise and cognition in both cross sectional and longitudinal samples. Qualitative and quantitat ive reviews of the literature have postulated that late life exercise results in cognitive benefits. Most reviews suggest that executive functions seem to be differentially benefited by exercise in older individuals (Colcombe, Kramer, McAuley, Erickson, & Scalf, 2004; Colcombe & Kramer, 2003; Hillman, Erickson, & Kramer, 2008; Kramer & Erickson, 2007; Kramer, Erickson, & Colcombe, 2006; McAuley, Kramer, & Colcombe, 2004) although general cognitive abilities profit as well. A meta analysis of randomized con trolled trials of exercise revealed that regular involvement in exercise improves reaction time, spatial processing, controlled processes, and executive functions in older adults (Colcombe & Kramer, 2003) Further,

PAGE 13

13 a longitudinal study reported that indivi duals with low baseline cardiorespiratory fitness (CRF) scores displayed greater decrease in global cognition over six years, while cognitive performance remained stable for those whose CRF was higher at baseline (Barnes, Yaffe, Satariano, & Tager, 2003) Similarly, exercise intervention studies have reported that even low intensity physical activity and aerobic exercise improve episodic memory (Ruscheweyh et al., in press) and executive functions respectively (Colcombe, Kramer, Erickson et al., 2004) Thus several research studies have identified a positive relationship between exercise and cognitive improvements in late life. Exercise and Brain Function in Late Life Functional magnetic resonance imaging (fMRI) studies have also uncovered changes in older Flanker task using fMRI before and after a six month (walking) intervention. The authors found that older ad ults in the exercise group showed faster reaction times and increased accuracy on the Flanker task when compared to those in the control group (stretching and toning). These changes in cognitive function were accompanied by concomitant brain activity chang es in attentional networks, suggesting that those who exercised displayed improved capacity to allocate attentional resources during the Flanker task (Colcombe, Kramer, Erickson et al., 2004) Similarly, long term engagement in physical activity has been shown beneficial for brain activity and psychomotor speed even two years after termination of a physical activity intervention in older adults (Rosano et al., 2010) Further, studies have reported that the functional brain connectivity of structures usuall y associated with attentional processes (posterior cingulate cortex and middle frontal gyrus), is a significant source of variance in the

PAGE 14

14 relationship between aerobic fitness and executive functions, and could thus be considered a plausible underlying mech anism of the exercise induced cognitive benefits observed in late life (Voss et al., 2010) Similarly, following a four month exercise intervention, Burdette and colleagues found heightened connectivity of the hippocampus to the anterior cingulate cortex, as well as increased cerebral bloo d flow to the hippocampus only in older adults assigned to an exercise group (Burdette et al., 2010) These functional brain imaging findings further support the potential of exercise to increase plasticity in late life by exerting changes at the neural level. Exercise and Brain Structure in Late Life Nevertheless, exercise has not only been associated with improved cognition and changes in functional brain activity and connectivity, but also with increases in gray and whit e matter brain volumes in late life. With age, older adults display reductions in gray and white matter volumes (Good et al., 2001; Raz & Rodrigue, 2006; Tisserand et al., 2004). The most pronounced reductions in gray matter occur in prefrontal, temporal, and parietal cortices (Raz & Rodrigue, 2006; Resnick, Pham, Kraut, Zonderman, & Davatzikos, 2003; Tisserand et al., 2004), while changes in white matter microstructure suggest that the anterior white matter is differentially affected by age (Raz & Rodrigue 2006). In order to measure the relationship between exercise and brain volume, some investigators have conducted structural brain studies in older adults who are aerobically fit or sedentary or who become aerobically fit following an exercise intervent ion. Colcombe and colleagues, in a cross sectional study, predicted that brain areas most affected by aging (frontal, temporal, and parietal lobes) would be spared of atrophy in older adults who display high aerobic fitness levels. The authors found that, indeed,

PAGE 15

15 older adults with higher fitness levels, displayed greater gray matter volumes in prefrontal, temporal, and superior parietal cortex, as measured by voxel based morphometry (VBM), than those with lower fitness scores (Colcombe et al., 2003) Furthe r, they found that the anterior corpus callosum, which is critical for inter hemispheric communication, was spared in older adults with higher fitness levels. The same findings were supported following a six month exercise intervention in which the aerobic training group displayed increased gray matter volumes in frontal and temporal brain regions, as well as increased white matter volumes in the anterior corpus callosum (Colcombe et al., 2006) Other cross sectional studies have reported associations betwe en aerobic fitness, hippocampal volume, and spatial memory ( Erickson et al., 2009) Longitudinal studies have supported the cross sectional literature indicating that physical activity at baseline predicts higher gray matter volume over nine years (Erickso n et al., 2010) Furthermore, exercise intervention studies have found positive effects of exercise on gray matter volume and cognitive function in older adults (Colcombe et al., 2006; Erickson et al., 2011; Ruscheweyh et al., in press) Given that gray m atter volumes have been associated with improved cognitive function in older adults, investigators have suggested that exercise induced neural changes may be the underlying mechanism by which exercise improves cognitive function (Colcombe et al., 2006; Col combe, Kramer, Erickson et a l., 2004; Erickson et al., 2009; Erickson et al., 2011) Animal studies have confirmed human findings by demonstrating improved water maze performance in rats and mice following exercise. These cognitive changes are accompanied by increases in angiogenesis, synaptogenesis, neurogenesis, and neurotrophic factors related to central nervous

PAGE 16

16 system plasticity, such as brain derived neurotrophic factor [BDNF] (Berchtold, Chinn, Chou, Kesslak, & Cotman, 2005; Pereira et al., 2007; Swai n et al., 2003; Vaynman & Gomez Pinilla, 2005; Vaynman, Ying, & Gomez Pinilla, 2004) Since older adults display deficits in executive functioning, and the frontal lobes are involved in these functions, increased cortical frontal volumes should result in i mproved executive abilities. Most structural brain studies of exercise in aging have not collected cognitive performance measures (Colcombe et al., 2003; Colcombe et al., 2006) or have simply described them in parallel to brain structure, without directly correlating brain volume differences with cognitive performance (Gordon et al., 2008) For example, Gordon and colleagues (2008) conducted a cross sectional study of brain str ucture and cognitive aging which aimed at further describing the relationship between aging, cardiopulmonary fitness, and education levels. They collected several neuropsychological measures; however, these measures were correlated to fitness and education levels separately and their association with brain structure was not directly assessed. Thus defining the relationship between gray matter volume and cognitive performance in older adults remains understudied. Similarly, the type and intensity of exercis e or physical activity needed to achieve neural and cognitive benefits is not well characterized in the literature. In an effort to better understand these differences, Ruscheweyh and colleagues (in press) investigated the intensity of exercise needed for observable cognitive effects in older adults. The authors studied the influence of different physical activity regimens on episodic memory function, gray matter volume, and neurotrophin levels. They found that physical activity as a whole (low intensity ae robic, medium intensity aerobic, and

PAGE 17

17 physical activity embedded within activities of daily living) can confer beneficial cognitive effects. On the other hand, one study suggested that levels of physical activity, and not cardiovascular fitness, were associ ated with improved memory performance, and physical activity was also related to higher gray matter volume in prefrontal and cingulate cortex (Flel et al., 2010) In sum, both cross sectional and exercise intervention studies have identified a strong lin k between aerobic fitness, physical activity and improved cognitive function in older adults. These changes in cognitive function seem to be supported by preservation or increases in gray and white matter volumes in areas critical for higher order cognitiv e functioning such as the frontal, temporal, and parietal cortex. The Present Study The purpose of the present study is to better characterize the differential effects of self reported physical activity and fitness level on gray matter volume and cognitiv e performance in older adults For this purpose, we collected structural magnetic resonance images (MRI) in a sample of neurologically healthy older adults in order to conduct voxel based morphometry (VBM) analyses. We also obtained executive function p erf ormance measures on the same sample to evaluate the relationship between fitness, physical activity, brain volume and cognitive function In addition, this study used cognitive and structural brain data from an independent exercise intervention study in an attempt to add converging evidence to our cross sectional findings. For this purpose, cognitive and structural brain MRI data from the Lifestyle Interventions and Independence for Elders pilot study (LIFE P) was analyzed (Pahor et al., 2006)

PAGE 18

18 CHAPTER 2 AIMS & HYPOTHESES Aim 1 To investigate the unique relationship between fitness levels, age, executive function, and gray matter volume in older adults, as well as the moderating role of fitness in the relationship between age and gray matter volume. Hypothesis 1 Fitness and executive function performance will add unique variance to the prediction of gray matter volumes, with higher fitness and higher executive functions associated with higher gray matter volumes. Age, however, will be negatively correlated to gray matter, especially within the prefrontal, temporal, and parietal cortex. Further, fitness level will moderate the relationship between age and gray matter volume. Aim 2 To investigate the unique relationship between self reported physical activity level s, age, executive function, and gray matter volume in older adults, as well as the moderating role of physical activity in the relationship between age and gray matter volume. Hypothesis 2 Self reported physical activity levels and executive function per formance will add unique variance to the prediction of gray matter volumes, with higher physical activity levels and higher executive functions associated with higher gray matter volumes. Age, on the other hand, will display negative associations with gray matter volume, especially in areas typically affected by biological aging such as prefrontal, temporal, and parietal

PAGE 19

19 cortex. Physical activity will moderate the relationship between gray matter volume and age based on findings from the literature. Aim 3 To examine whether gray matter volume predicts executive function performance above and beyond fitness/physical activity and age. Hypothesis 3 Gray matter volume will predict executive function performance above and beyond the contribution of fitness/phy sical activity and age. Aim 4 To use the LIFE P data to compare the cognitive and gray matter volume differences between participants assigned to the Physical Activity (PA) group and those assigned to the Successful Aging control group (SA) two years aft er termination of the physical activity intervention. Hypothesis 4 Participants from the LIFE P study who were previously assigned to the Physical Activity (PA) group will display higher gray matter volumes than those assigned to the Successful Aging con trol group (SA), especially within areas vulnerable to biological brain aging (frontal, temporal, parietal cortex). Participants in the PA group will also display better performance on a task of executive function when compared to the SA group.

PAGE 20

20 CHA PTER 3 C ROSS SECTIONAL STUDY DATA Methods Participants Three groups of participants were originally collected as part of a larger study: physically active older adults, sedentary older adults, and a young adult group. For the purposes of the current study, and given the overlap of physical activity and fitness measures between the physically active and sedentary older groups, we utilized the data from older adults only, with physical activity and fitness as continuous variables. The original physica lly acti ve group consisted of 16 neurologically healthy older adults (ages 60 85) who self reportedly performed voluntary aerobic exercise lasting at least 30 minutes for a minimum of 3 times per week. The original sedentary older a dult group was matched for age to the physically activ e group, and was comprised of 13 neurologically healthy older adults, ages 6 1 8 1, who by self report, performed aerobic exercise less than 45 minutes per week. Data from 15 h ealth y young adults (ages 18 30) was also collected, but it is not included in the present study. Thus the total sample size for the current study is 29 older adults. All participants were healthy community dwelling older adults recruited from the general Gainesville FL area. Signed informed consent was obtained from all participants according to guidelines established by the Participants were compensated for their participation in the study. See Table 3 1 for participant demographics

PAGE 21

21 Exclusion Criteria All participants were right handed, native English speakers, who were deemed eligible for MRI scanning following an extensive screening protocol (e.g., no cardiac pacemaker, ferrous metal implants, or claustrophobia). Participants w ere free of a disease, ild c ognitive i mpairment), history of head trauma with loss of consciousness, cardiac conditions, learning disabilities, attention defici t disorder, history of alcohol or drug abuse (for at least six months prior to participation), and psychiatric conditions. Older adults currently prescribed beta blockers for hypertension management were not included in the study. On neuropsychological tes ting, all subjects obtained Mini (Folstein, Folstein, & McHugh, 1975) indicating no global cognitive impairment Aerobic Fitness and Physical Activity Assessment A modified Cooper treadmill test was conducted (Co oper, 1968) in which participants were asked to cover as much ground as possible by walking, jogging, or running on a treadmill for twelve minutes. Distance traveled in twelve minutes (self paced, variable speed) was the main fitness outcome measure. To me asure self reported physical activity level, we used a modified version of the Leisure Time Exercise Questionnaire (LTEQ), which is a three item scale that asks participants to rate how often they engaged in mild, moderate, and strenuous leisure time exerc ise ( Godi n, Jobin, & Bouillon, 1986; Godin & Shephard, 1985) The LTEQ is a reliable and valid measure of leisure time exercise behavior in adults (Jacobs, Ainsworth, Hartman, & Leon, 1993) that allows for calculation of total number of minutes spent in mo derate and strenuous physical activity which was modified in the present study to obtain total

PAGE 22

22 minutes of moderate and strenuous physical activity performed in one week. Hence, d istance travelled during the treadmill test was considered a measure of fitne ss level, while minutes spent performing moderate and strenuous physical activity, as measured by the modified LTEQ assessed subjective physical activity level. Neuropsychological Assessment Each participant completed approximately two hours of neuropsych ological testing for the larger study For the purposes of this study, we created an executive function composite score for each participant given the reported declines in executive function that occur as a function of age, as well as the beneficial effect s that exercise has been suggested to exert on measures of executive function in older adults. The executive function composite score was comprised of the averaged total scores for the following measures: letter fluency total correct (FAS), number correct on a reasoning measure (Letter Series*), Trails B score, reaction time (RT) during a modified version of the Flanker Task*, Lexical Decision task* RT, and Number Copy and Symbol Digit* RT. Measures marked with an asterisk (*) were computerized for ease of administration and to obtain RT measures as well as accuracy. Computerized measures were programmed using DirectRT v2004 and MediaLab experimental research software (Jarvis, 2004a, 2004b) All cognitive variables were analyzed using PASW statistics 18, Rel ease version 18.0.0 (PASW/SPSS, 2009) Structural Imaging Structural MRI images were acquired on a 3 Tesla Achieva whole body scanner (Philips), with an 8 channel SENSE radio frequency head coil, at the McKnight Brain Institute of the University of Florida. For the VBM analyses, structural TFE T1 weighted images were acquired for 160 x1mm sagittal slices (FOV=240mm ; TE =3.685 ms;

PAGE 23

23 TR=8.057 ms; FA=8 degrees; matrix=256x256, voxel size=1.0mm x 0.938mm x 0.938mm). Procedure An exercise screening questionnaire and an MRI screen were conducted via telephone to identify potential study candidates. For the larger study, older ad ults who reportedly performed voluntary aerobic exercise lasting at least 30 minutes for a minimum of 3 times per week were temporarily assigned to the physically active group. Those older adults who reportedly performed aerobic exercise less than 45 minut es per week were temporarily assigned to the sedentary older adult group. Given the nature of the current study, we did not analyze data by the assigned groups, but rather used fitness and physical activity as continuous variables. Older adults who qualifi ed for the study based on the MRI and exercise phone screens, were asked to obtain written clearance from their primary care physician in order to participate in the treadmill test. he first study session. During the first session, which lasted approximately 2.5 hours, participants underwent neuropsychological testing first (2 hours), followed by the aerobic fitness treadmill test (12 minutes). Heart rate was monitored throughout the treadmill assessment. Participants were then instructed on how to answer the modified LTEQ, which they completed for a monitoring period of seven days following the first session. Structural imaging scans were obtained during the second session, which was scheduled following the seven day monitoring period. At this time, participants brought the completed LTEQ to assess self reported physical activity levels.

PAGE 24

24 Data Analyses Structural Imaging Image preprocessing MRI structural images were analyzed using FS L VBM, a voxel based morphometry (VBM) style analysis (Ashburner & Friston, 2000; Good et al., 2001) carried out with FMRIB Software Library (FSL) version 4.1 tools ( Smith et al., 2004) This technique allows for the identification of different types of brain tissue (i.e., gray matter, white matter, and cerebrospinal fluid or CSF) at the voxel level by calculating the probability that each voxel contains a particular tissue type. Pre pr ocessing of the structural images for VBM analysis consisted of the following steps: a) removing the skull and other non brain tissue from the structural images using flsvbm_1_bet ( Smith, 2002) b) performing tissue type segmentation to generate partial gr ay matter volumes for each participant in the native space using FAST4 (Zhang, Brady, & Smith, 2001) c) aligning the resulting gray matter partial volumes to MNI152 standard space using the non linear registration tool FNIRT (Andersson, Jenkinson, & Smith 2007a, 2007b) which uses a b spline representation of the registration warp field (Rueckert et al., 1999) The resulting images were concatenated and averaged to create a study specific template containing the registered gray matter images for each of t he 29 older adult subjects (fslvbm_2_template n). All of the gray matter images were then non linearly registered to the study specific template (fslvbm_3_proc). To account for changes in voxel size that occur within the image registration process (expans ion and contraction), all registered partial volume images were modulated by dividing by the Jacobian of the warp field (fslvbm_3_proc). The modulated, segmented images were then smoothed

PAGE 25

25 with an isotropic Gaussian kernel of sigma 2mm, which corresponds to 4.6mm full width half max (FWHM). Image analyses: regression models All predictor variables entered into the model were first normalized using the Blom transform method available on PASW statistics 18, Release version 18.0.0 (PASW/SPSS, 2009) software t o meet GLM assumptions. Fitness level and executive functions composite were orthogonalized (removed shared variance) due to their significant correlations with age (fitness r = .40, p =.03; executive r = .47, p =.01) to avoid problems resulting from multico llinearity (Pedhazur, 1997) For this purpose, a simple regression model with fitness as the dependent variable and age as the predictor was conducted while saving the unstandardized residuals. This procedure results in a residualized fitness variable (fit ness age ) from which the shared variance between age and fitness has been removed and represents the unique variance associated with fitness level. The same procedure was followed for the executive function composite, deriving an executive composite variab le that is independent of age (executive age ). Subsequently, age, physical activity, and the interaction terms were centered (mean of 0) as required for FSL tools to carry out multiple regression analyses. We used the randomize option in FSL which is a permutation based program allowing for modeling and inference testing using standard general linear model design, when the null distribution of a statistic map is unknown. To describe the unique relationship between gray matter, fitness level, physical act ivity level, and cognitive function, we conducted two voxel wise multiple regression models under the general linear model, higher level/non time series design option in FSL. The first regression model used voxel wise gray matter volume as the dependent va riable, and fitness age

PAGE 26

26 (distance travelled in 12 minutes during treadmill test), age, executive age function composite, and the interaction between age and fitness age (age X fitness age ) as predictors. The second regression model included voxel wise gray matter volume as the dependent variable and physical activity (self reported minutes of moderate + strenuous activity in one week), age, executive age function composite, and the interaction between age and physical activity (age X pa) as predictors. The resulting probability maps were corrected for family wise error ( p <.05) using the T hreshold F ree C luster E nhancement (TFCE) method implemented on FSL. T his method enhances cluster like regions more than background (noise) by creating output that is a weigh ted sum of all of the local clustered signal, without the need for arbitrary cluster thresholding (Smith & Nichols, 2009). To obtain parameter estimates as indicators of effect size, regardless of significance level, we performed another set of ordinary le ast squares regressions based on the same models described above using FSL FEAT. The resulting statistic maps provided unstandardized b weights (b) for each of the predictors entered into the model. Each b weight statistic map was transformed to standardiz ed beta independent variable/standard deviation of the dependent variable). Neuropsychological Data ll cognitive variables were normalized using the Blom transform method available within PASW statistics 18, Release version 18.0.0 (PASW/SPSS, 2009) After normalization, all RT measures were reflected so that higher values represent better performance. Then, the Blom transformed scores of all cognitive measures of executive function (as described in the methods section) were averaged to create the exe cutive function composite score

PAGE 27

27 for each participant, which was entered into the regression analyses as a predictor of gray matter volume on each of the two regression models carried out on FSL. To examine whether gray matter volume predicted executive fu nction performance above and beyond the contribution of fitness or physical activity and age, hierarchical regression models were carried out on PASW statistics using mean gray matter volumes extracted from specific regions of interest (see results section ). Results Fitness Regression in FSL Fitness age age, and executive age function did not significantly predict gray matter volume in any brain region after correcting for multiple comparisons ( TFCE p <.05) The interaction between age and fitness age h owever, survived family wise correction at the p <.05 level. Areas that survived family wise error correction include: left and right cerebellum, left hippocampus and parahippocampal gyrus, and right and left inferior lateral occipital cortex/occipital fusi form, and left lingual gyrus. See Figure 3 1 for a depiction of the brain regions in which the interaction between age and fitness age predicted gray matter volume. Fitness Regressions in PASW To decompose the significant interaction and better understand the relationship between age, fitness and gray matter volume in these regions, a mask of the significant areas was created and separated into the five distinct anatomical regions it comprised (left cerebellum, right cerebellum, left hippocampus/parahippoca mpal gyrus, right occipital fusiform/in ferior lateral occipital cortex left occipital fusiform/inferior lateral were extracted from these regions and exported from FSL to be analyzed with PASW

PAGE 28

28 software. Multiple regression models with fitness age age, executive age function, and age X fitness age predicting gray matter volume at each of the five ROIs were computed. The fitness age variable was dichotomized (see Table 3 2 f or participant characteristics per fitness group) using a median split in order to visually plot the interactions between fitness, age, and gray matter volume at each of these regions. See Figure 3 3 for scatter plots depicting the interaction between age and fitness age level on gray matter volume at each of the ROIs. For the left cerebellum (Lcer), the regression model was significant, F (4,24)=9.4, p <.001, Adj R 2 =.55. The interaction between age and fitness age was significantly associated with gray matte r volume, where participants with lower fitness levels did not show the beneficial effects of fitness on gray matter volume as age increased: age X fitness t (24 )= 5.82 p <.0 01 .76. Within the right cerebellum (Rcer), the overall model was significant F ( 4,24)=7.78, p <.001, Adj R 2 =.49. Fitness level significantly moderated the relationship between age and gray matter volume, where higher fitness level had a higher impact on gray matter volume with increasing age: age X fitness t (24 )= 5.42 p <.0 01 .75. Th e regression model was significant for the left hippocampus/parahippocampal gyrus (Lhippo) as well F (4,24)=11.91, p <.001, Adj R 2 =.61. Fitness age age, and the interaction between age and fitness age level were significantly associated with gray matter vol ume. Here, those individuals with lower fitness scores did not display the mitigating effect of fitness on gray matter volumes as age increased when compared to those with higher fitness levels ; fitness t (24 )= 3.14 p <.0 1 .38; age t (24 )= 3.77 p <.01 = .45; age X fitness t (24 )= 5.58 p <.0 01 .67. For the right occipital fusiform/inferior lateral occipital cortex ROI (RLOC), the

PAGE 29

29 regression model was significant F (4,24)=10.69, p <.001, Adj R 2 =.58. Fitness age level and the age X fitness age interaction w ere significantly associated with gray matter volume in this area, where those with higher fitness scores showed increased impact of fitness on gray matter volume as age increased: fitness t (24 )= 2.71 p <.0 5 .34; age X fitness t (24 )= 6.38 p <.001 .80. Finally, for the left occipital fusiform/inferior lateral occipital cortex/lingual gyrus (LLOC) ROI, the regression model was significant F (4,24)=13.67, p <.001, Adj R 2 =.64. Fitness age age, and age X fitness age were significantly correlated to gray matter volume in this region, in which those with lower fitness scores did not display a protective effect of fitness on gray matter as age increased: fitness t (24 )= 2.56 p <.0 5 .30; age t (24 )= 3.94 p <.01 .45; age X fitness t (24 )= 6.25 p <.0 01 72. Refer to Figure 6 1a for a graphed representation of the median s at each of the ROIs. The positive slope of the relationship between age and gray matter volume in the Rcer and RLOC within the high fitness group was not due to the older individuals in this group being more educated than those in the low fitness group, since age was not significantly correlated to education within the high ( r = .16, p =.60) or low fitness groups ( r =.20, p =.49). Physical Activity Regression in FSL Physical activity, age, executive function age and the interaction between age and physical activity did not significantly predict gray matter volume after correcting for multiple comparisons ( TFCE p <.05) Direct Comparison between Fitness age & Physical Activity in FSL Based on the above results, it would appear that the age X fitness age interaction is a better predictor of gray matter volume than self reported physical activity or its associated interaction with age. However, a head to head comparison of these

PAGE 30

30 potential pre dictors is necessary to make such a statement definitively. To confirm if the age X fitness age interaction was a better predictor of gray matter volume than physical activity level, a different whole brain multiple regression model was conducted on FSL w ith fitness age physical activity, age X fitness age and age X physical activity as predictors of voxel wise gray matter volume. Results indicated that the interaction between age and fitness age was the only significant predictor of gray matter volume ( TFCE p <.05) The areas in which fitness moderated the relationship between age and gray matter volume in the current analysis were much smaller than, but encompassed by the same anatomical regions as the significant areas in which the age X fitness age interaction predicted gray matter volume in the previous regression analysis (refer to Fitness Regression in FSL section). None of the other potential predictor variables (f itness age alone, physical activity alone, age X fitness age age X physical activi ty) predicted gray matter volume in any brain regions after correcting for multiple comparisons (TFCE p <.05). Thus the interaction between age and fitness age was a better predictor of gray matter volume than physical activity level or its associated inte raction with age in a head to head comparison. See Figure 3 2 for the overlap of significant regions between the two regression models. Neuropsychological Data Two regression models were conducted to predict executive function in the current sample. The f irst was a hierarchical regression model in which the executive function composite was the dependent variable. The predictors were entered in the following order: block 1=fitness age ; block 2=age; block 3=age X fitness age ; and block 4= mean gray matter vo lume for the average of the five ROIs (Lcer, Rcer, Lhippo, RLOC, LLOC). Fitness age alone did not significantly predict executive function, F (1,27 )= .12 p = 73,

PAGE 31

31 R 2 =. 005, Adj R 2 = .032 Adding age to the model significantly increased variance explained by 22% ( p =.012) above and beyond that explained by fitness age alone, F (2,26 )= 3.71 p < 05, R 2 =. 22, Adj R 2 =.16. When adding the interaction term between age and fitness age into the model, there is a significant increase in variance explained by 12% ( p =.042), F (3,25 )= 4.34, p < 05, R 2 =. 34, Adj R 2 =.26. The addition of gray matter volume to the model did not significantly contribute additional variance in executive function scores, F (4,24 )= 3.13 p < 05, R 2 =. 34, Adj R 2 =.23, R 2 change=0%, p =.92 Coefficients for the last model indicate that age is the only predictor that contributes uniquely to the total variance in executive functions; fitness age t ( 24)= .76, p =.46 .15; age t ( 24)= 2.72, p <.0 5 .49 age X fitness age t ( 24 )= 1.18 p = 25 .38; mean gray matter volume t ( 24 )= .11 p =.92 .04. The second hierarchical regression model included the following predictors of executive function: block 1=physical activity; block 2=age; block 3=age X physical activity; and block 4= mean gray matter volume for the average of the five ROIs. Physical activity alone did not predict executive function, F (1,27 )= 1.85 p = 19, R 2 =. 064, Adj R 2 =.029. Adding age to the model contributed 23% additional variance ( p <.01), F (2,26 )= 5.52 p <.05, R 2 =. 30, Adj R 2 =.24. The addition of the age X physical activity interaction did not contribute additional variance ( R 2 change=0%, p =.92) to the model, F (3,25 )= 3.54 p < 05, R 2 =. 30, Adj R 2 =.21. Similarly, gray matter volume did not significantly explain any additional variance ( R 2 change=4%, p =.25) on executive function scores, F (4,24 )= 3.06 p < 05, R 2 =. 34, Adj R 2 =.23. Coefficients for the last model indicate that only age contributes uniquely to the total variance in executive fun ctions; physical activity t ( 24)= 1.10, p =.29 .2; age t ( 24)= 2.47, p <.0 5 .43; age X pa

PAGE 32

32 t ( 24 )= .28 p = 79 .05; mean gray matter volume t ( 24 )= 1.19 p =.25 .22. See Appendix for a graphed representation of median s by ROI (Figure 6 1b). Discussion It has been consistently shown in the literature that biological aging is accompanied by neural challenges and concomitant declines in cognitive performance. Several studies have suggested that exercise and increased physical activity in older a dults can help mitigate the effects of biological brain aging, thus delaying or preventing cognitive decline. There are discrepancies in the literature regarding the differential contribution of aerobic fitness and physical activity to the preservation of brain tissue and cognitive performance in late life. Moreover, many studies assessing the relationship between brain volume and fitness levels have not measured the direct and unique relationship between cognitive function and brain volume, but rather asse ss it indirectly or not at all (Colcombe et al., 2003; Colcombe et al., 2006; Gordon et al., 2008) The main purpose of the current study was to evaluate whether fitness and/or physical activity better predict gray matter volume and if they moderate the re lationship between age and gray matter volume, while also assessing the direct relationship between brain volume and executive function in older adults. We found that the interaction between fitness level and age predicted gray matter volume in bilateral occipital and cerebellar regions, as well as in the left hippocampus, while self reported physical activity level did not predict gray matter volume in any brain region in this sample. Thus, the interaction between age and fitness level was a better predi ctor of gray matter volume than self reported physical activity level in the current sample. This finding is not surprising given the self report nature of the physical activity variable, which is confounded by participants accurately assessing and recalli ng the

PAGE 33

33 type and frequency of activities they have performed daily, for seven days. Further, this self assessment was a snap shot of only one week, which may not have been representative of most weeks for all participants. Our measure of fitness, on the ot her hand, was an objective measure of distance travelled on a treadmill over a 12 minute period and it was a modification of a well established method to derive aerobic fitness levels (Cooper, 1968) Although this fitness measure is not optimal, and assess ing maximal oxygen consumption (VO2 max ) would have been preferred, we were able to find that fitness level significantly moderated the relationship between gray matter volume and aging in five distinct brain regions: the left cerebellum, right cerebellum, right occipital fusiform/inferior lateral occipital cortex, the left hippocampus and parahippocampal gyrus, and the left occipital fusiform/inferior lateral occipital cortex/lingual gyrus. The hippocampus, cerebellum, and occipital cortices have been shown to decline as a function of age. The hippocampus has been shown to shrink at a medium rate of 1.23% each year. Similarly, the median correlation of brain volume with age in the cerebellum is r = .30, and in occipital cortex is estimated at r = 0.19, (Raz & Rodrigue, 2006) Thus, it is not surprising that higher fitness levels would help mitigate the effects of age on brain volume in these regions. Moreover, positive associations between fitness level and brain volume in these regions have been previously re ported in the literature (Erickson et al., 2009; Erickson et al., 2010; Erickson et al., 2011; Honea et al., 2009) However, we were unable to find associations between fitness and gray matter volume in frontal, lateral temporal, and parietal cortices as p reviously reported in the literature (Colcombe et al., 2003; Colcombe et al., 2006; Erickson et al., 2010; Flel et al., 2010; Ruscheweyh et al., in press) Given that these regions are

PAGE 34

34 amongst the first ones to be affected by aging, and fitness level has been shown to reduce tissue brain loss in the same areas (Colcombe et al., 2003; Colcombe et al., 2006) it is unclear why we were unable to find such associations in the current sample. Previous studies do not report whether they orthogonalized their pred ictor variables before entering them into the regression analyses (Colcombe, 2003). If fitness level is not made independent from age in a regression analysis, it is possible that the shared variance between age and fitness might explain most of the varian ce associated with gray matter volume in the frontal and temporal lobes. This could potentially explain why we obtained significant results in occipital rather than frontal brain regions, given that our fitness variable was independent from age and represe nted the unique effect of fitness on gray matter volume rather than the shared effect of fitness and age on gray matter volume. Furthermore, given that our measure of fitness is different from those usually reported in the literature (i.e. VO2 max or estima tes thereof), the current study may have captured variance associated with a different aspect of fitness level that does not overlap completely with measures such as VO2 max ( which assesses cardiac output ) Another likely explanation for the current findings is that power to detect significant differences was affected given our small sample size (N=29). The same can be said regarding the lack of contribution of the executive composite measure to predicting g ray matter volume, as well as the lack of contribution of gray matter volume as a unique predictor of executive function. First, the small sample size makes it difficult to find significant associations. Second, the executive function composite used in the current study differs from the typical measures reported in the VBM and exercise literature (i.e., spatial, episodic, and verbal memory, intelligence tests) Similarly, some of the tests

PAGE 35

35 included in the executive function composite (i.e., number copy and symbol digit ) may be considered a measure of psychomotor speed rather than executive functions by some investigators, possibly extending the interpretation of the composite variable to an area other than executive function. This circumstance is further com plicated by the fact that detecting associations between brain volume and cognitive measures has proven states, the nature of the measures used to assess cognitive function, and the composition of the sample (Raz, 2000; Raz & Rodrigue, 2006) In sum, f itness level moderated the relationship between gray matter volume and age such that the impact of fitness level became more important as age increased. Given that g ray matter volume decreases with age, it should be expected that the impact of fitness level becomes more important as age increases and neural challenges become more apparent. Hence, remaining fit in late life seems more impactful to neuroplasticity once neural challenges, such as gray matter volume loss, have taken place. The current findings attest to the neuroprotective role that fitness plays in late life, by sparing gray matter volume loss as age increases. Executive function performance was not a si gnificant predictor of gray matter volume in the current study and gray matter volume did not predict executive function performance above and beyond the contributions of age. Future research investigating the relationship between fitness, physical activit y, cognitive function, and brain volume should be longitudinal/interventional in nature and measure physical activity and fitness levels utilizing objective variables rather than self report measures. Similarly, cognitive performance should be measured con sistently between studies to reduce

PAGE 36

36 methodological variability and increase scientific comparability. Furthermore, future research should report on the correlation between predictor variables (i.e. fitness, age, cognitive measures) to better understand the unique contributions of each variable to gray matter volume.

PAGE 37

37 3 1 Cross s ectional study p articipan t c haracteristics Mean SD Minimum Maximum Age 68.38 5.98 60.00 85.00 Education 16.14 2.36 12.00 20.00 Gender 18 female / 11 male PA 242.86 204.78 0.00 900.00 Fitness 0.55 0.21 0.19 1.01 MMSE 28.97 0.98 27.00 30.00 Executive 0.00 0.69 1.19 1.14 Lcer 0.61 0.12 0.38 0.81 Rcer 0.59 0.12 0.34 0.84 Lhippo 0.35 0.07 0.22 0.50 RLOC 0.66 0.14 0.44 0.98 LLOC 0.65 0.15 0.39 0.96 Notes: SD = standard deviation; PA = physical activity variable; MMSE = Mini Mental State Examination; Executive = Blom transformed executive function composite; the remaining variables depict the mean gray matter volume for each of the following regions o f interest (ROI): Lcer = left cerebellum; Rcer = right cerebellum; Lhippo = left hippocampus/parahippocampal gyrus; RLOC = right inferior lateral occipital cortex/occipital fusiform gyrus; LLOC = left inferior lateral occipital cortex/occipital fusiform gy rus/lingual gyrus. 3 2 Cross s ectional study participant characteristics by median s plit Median Split of Fitness Variable Low Fitness (N=15) High Fitness (N=14) Mean SD Min Max Mean SD Min Max Age 67.20 5.56 60 81 69.64 6.37 60 85 Education 15.67 2.77 12 20 16.64 1.78 14 20 Gender 12 female 6 female PA 156.19 142.39 0 460 335.71 224.80 120 900 Fitness 0.42 0.17 0.19 0.66 0.68 0.16 0.42 1.01 MMSE 28.87 1.13 27 30 29.07 0.83 27 30 Executive 0.08 0.66 1.05 1.14 0.09 0.74 1.19 1.02 Notes: SD = standard deviation; PA = physical activity variable; MMSE = Mini Mental State Examination; Executive = Blom transformed executive function composite

PAGE 38

38 3 1 Age X fitness i nteraction: Significant c lusters Red/orange regions depict the areas in which fitness age moderated the relationship between age and gray matter volume. These regions correspond to the right (Rcer) and left cerebellum (Lcer), left hippocampus/parahippocampal gyrus (Lhippo), and right inferior lateral occipital cortex/occipital fusiform gyrus (RLOC) and left inferior lateral occipital cortex/occipital fusiform gyrus/lingual gyrus (LLOC).

PAGE 39

39 3 2 Age X f itness i nteraction : O verlap b etween two regression m odels The areas in blue represent clusters in which the interaction between age and fitness age was significant (TFCE p <.05) from a multiple regression model in which fitness age physical activity, age X fitness age and age X physical activity predicted voxel wise gray matter volume. The regions in orange represent the si gnificant interaction between age and fitness age derived from the multiple regression model in which fitness age age, executive function, and age X fitness age predicted voxel wise gray matter volume. It can be seen that the areas in blue are much smalle r than those in orange, but are encompass ed by the same anatomical regions as the significant areas from the previous analysis. T hus the interaction between age and fitness age was a better predictor of gray matter volume than self reported physical activi ty and its associated interaction with age

PAGE 40

40 3 3 Age X fitness i nteraction: Scatter plots Y axis depicts mean gray matter volume extracted from each of the five ROIs resulting from the age X fitness age interaction (Figure 3 1). X axis represents Blom transformed age, where 2 = 60; 1 = 61; 0 = 68; 1 = 74; and 2 = 85. High and low fitness lines are based on a median split for the fitness age variable. Lcer = left cerebellum; Rcer = right cerebellum; LLOC = left inferior lateral occipital cortex/occipit al fusiform gyrus/lingual gyrus; RLOC = right inferior lateral occipital cortex/occipital fusiform gyrus; Lhippo = left hippocampus/parahippocampal gyrus. Indicates the linear trend is significant at the p <.05 level.

PAGE 41

41 CHAPTER 4 LIFE PILOT STUDY DATA Methods The LIFE P Study The Lifestyle Interventions and Independence for Elders Pilot study (LIFE P) was a single blinded randomized control trial of physical activity that lasted 1 year and was conducted at four different institutions: Cooper Inst itute, Stanford University, University of Pittsburgh, and Wake Forest University (Pahor et al., 2006) The intervention consisted of random assignment of sedentary older adults into two groups: a moderate intensity physical activity intervention (PA) or a health education successful aging (SA) group. The original trial collected data from 424 participants. Physical activity intervention (PA) The PA intervention consisted of a combination of aerobic, strength, balance, and flexibility exercises and it was divided into three phases: adoption (weeks 1 8), transition (weeks 9 24), and maintenance (weeks 25 to end of trial). During the adoption phase, three center based exercise sessions (40 60 minutes) per week were conducted within a supervised setting. The t ransition phase consisted of 4 months in which the number of center based sessions was reduced to 2 times/wk and home based maintenance phase, the intervention was home based, with optional once to twice per week center based sessions and monthly telephone contact. Walking was the primary mode of exercise with a goal of at least 150 minutes per week in which moderate intensity exercise was encouraged (13 on the Borg scale).

PAGE 42

42 Succe ssful aging intervention (SA) The SA intervention served as the active control group for the PA group and was designed to provide attention and health education. Participants met in small groups weekly for the first 26 weeks and then monthly. Topics of discussion included nutrition, me dications, foot care, and recommended preventative services at different ages. Basic educational information related to physical activity was provided. Each session ended with a short 5 10 minute intervention of gentle upper extremity exercise. LIFE P in clusion criteria Participants between the ages of 70 89 years who had a sedentary lifestyle at baseline (<20 minutes/wk of structured PA during the past month) were included in the LIFE P study. All participants had to be able to walk 400 m within 15 minut es without sitting or using an assistive device. A score of 9 (on a scale of 0 12) was needed in the Short Physical Performance Battery, as well as having completed behavioral tests that logged health behavior for one week. All participants gave informed c onsent, lived in the study area, and were not planning to move for at least 9 months. The LIFE P Follow up Brain Imaging Study This follow up study was conducted using a subsample of the LIFE P subjects from the Pittsburgh site. The details of the stu dy can be found elsewhere (Rosano et al., 2010) but the main aspects pertinent to the current study are presented here. At the Pittsburgh location, 104 participants completed the LIFE P trial. Two years after termination of the LIFE P study, participants were screened for interest in participation in a follow up functional magnetic resonance imaging (fMRI) study. Only participants who reported to have remained adherent to their original group assignment (PA or SA) after the LIFE P study were eligible for t he follow up fMRI study. To measure adherence

PAGE 43

43 to original group assignment, participants originally assigned to the PA group were asked: P s tudy ended, have you completely stopped your regular pa rticipants were inclu was asked P s tudy ended, have you spent at least 20 minutes a week getting regular exercise? Regular exercise includes activities like: brisk walking, jogging, weight lifting, cycling, to their original group assignment, current physical activity levels were measured via the Community Healthy Activities Mode l Program for Seniors (CHAMPS) questionnaire (Stewart, Verboncoeur, & McLellan, 2001) The CHAMPS measures the participation (yes or no) and frequency (days/wk and hours/day) for the following activities: walking leisurely, walk for errands, ride a bicycle use aerobic machines, exercise (strength and water exercise), stretching, and general conditioning. Mean kilocalorie expenditures were computed using the CHAMPS. Kilocalorie expenditures were computed from the following activities: walk leisurely, walk f or errands, ride a bicycle, use aerobic machines, do water exercise, stretching, heavy strength exercise, light strength exercise, and general conditioning. The main purpose of the follow up study was to compare the functional brain activity level of the PA and SA groups during an executive function task (Digit Symbol Substitution Test: DSST) two years after the LIFE P intervention ended, to assess the long term effects of adherence. It was found that participants in the PA engaged in more minutes of physi cal activity and had higher DSST scores than the SA group two years following the intervention. The PA group also displayed more brain activity in areas

PAGE 44

44 associated with processing speed, such as the left dorsolateral prefrontal cortex, posterior parietal, and anterior cingulate cortex (Rosano et al., 2010) Thirty participants who remained adherent to their original LIFE P assigned group were included in this follow up fMRI study (20 from the PA and 10 from the SA group). These participants underwent fMRI and a clinical exam. Age did not differ significantly between the PA and SA groups (80.8 & 81.45 respectively, p =.5) in the follow up subsample, although DSST accuracy and response time during the fMRI was significantly different between the groups (see Ro sano, et al. 2010 for details). The Current Brain Imaging Study The current study used a subsample of the LIFE P follow up fMRI study to compare the gray matter volume and cognitive performance between the PA and SA groups in an attempt to supplement ou r cross sectional findings. We hypothesized that the PA group would show higher levels of gray matter volume in frontal, temporal, and parietal areas when compared to the SA group, given the protective effect of exercise on brain health (Colcombe et al., 2 006) Participants The current study included data from 20 participants out of the 30 available from the LIFE P follow up fMRI study. Data were from 10 participants from the PA and 10 from the SA groups. Given that VBM analyses necessitate the creation of study specific templates, the sample size of each group must be equal in order for the study specific template to accurately represent each group. From the 20 structural images available in the PA group, 10 were selected by matching to the SA group by age The resulting groups did not differ significantly in age, calories expended in all exercise related activities, or score in the DSST, although the mean amount of calories expended in all

PAGE 45

45 exercise related activities and performance in the DSST was higher in the PA than SA group. See Table 4 1 for participant characteristics. Data Analyses Behavioral Analysis Independent samples T tests were carried out on PASW to compare the PA and SA groups on age, calories expended in all exercise related activities, and accuracy/reaction time on the DSST. Structural Images All MRI images were collected as described in Rosano, et al., 2010. A 3 Tesla Siemens Trio MR scanner with a 12 channel head coil was used. For the purposes of the current study, only T1 weighted i mages were used in order to perform the VBM analysis. T1 weighted images were three dimensional magnetization prepared rapid gradient echo (TR=2.3 seconds; TE=3.43 ms; slice thickness=1mm; 176 slices). All structural images were pre processed following the same procedures outlined on Chapter 3, data analyses section: structural images image preprocessing. The only difference is that the current study specific template was comprised of the 20 participants derived from this sample rather than the participan ts in the cross sectional sample. The resulting registered and modulated gray matter images were smoothed at 4.6 mm FWHM. The GLM higher level/non time series analysis option in FSL was used to carry out voxel wise group comparisons between the PA and SA groups using randomize, which corrects for multiple comparisons using a threshold free cluster enhancement method (TFCE) via permutation analyses. The following comparisons were modeled: PA SA & SA PA. Given our small sample size we also calculated voxe l d values to

PAGE 46

46 report effect sizes using the following formula on fslmaths: d =2t/df, where df is the degrees of freedom for the Ttest or 2n 2; where n=# of participants in each group (df=18). Results Group Comparisons: Demographics and Perform ance Mean age did not differ between the PA ( M =81.11, SD =3.05) and SA ( M =81.56, SD =2.77) groups, t (18)= .342, p d = .04. Calories expended in all exercise related activities at the time of follow up did not differ significantly between the PA ( M =2504.79, SD =1426.11) and SA ( M =1587.23, SD =905.83) groups, t (18)=1.72, p d =.19. Similarly, there was no significant difference between the PA ( M =81%, SD =15%) and SA ( M =66%, SD =27%) groups on accuracy during the DSST, t (18)=1.47, p >.05, Cohe d =.21. Reaction time during the DSST did not differ between the groups (PA: M =1836.79 ms, SD =356.84 ms; SA: M =2167.14, SD =507.12), t (18)= 1.68, p d = .19. Group Comparisons: Gray Matter Volume There were no gray matter regions that differe d significantly between the groups after correcting for multiple comparisons ( TFCE p <.05). See Appendix (Figure 6 2) for median effect sizes per ROI. Discussion We used data from an independent exercise intervention study in an attempt to verify our cross sectional findings and assess the relationship between physical activity level and brain density between groups previously assigned to a moderate intensity physical activity intervention (PA) versus a health education successful aging group (SA). We hypot hesized that participants previously assigned to the PA intervention

PAGE 47

47 group would display greater gray matter volume than those assigned to the SA group, especially in areas vulnerable to brain aging (e.g., frontal, temporal, parietal lobes). Results indica ted that there were no significant differences between the two groups on gray matter volume, nor did they differ significantly on their performance on an executive function measure (DSST) or calories expended in all exercise related activities. The fact that we did not find significant differences between the exercise and control groups is probably due to the small sample size included in this study, which affects the power needed to detect significant differences. Similarly, the lack of differences may h ave been due to the fact that physical activity level was measured via self reported adherence to a physical activity intervention that took place two years prior to the follow up fMRI study. It is possible that participants either over or underestimated t heir current level of physical activity and their adherence to the previous intervention. If this is the case, then the PA and SA groups may not be that different from each other and we should not expect to find significant differences in gray matter volum e. Moreover, incorporating a measure of fitness level rather than self reported physical report. Future intervention studies should measure the moderating effect of increased fitness level o n the relationship between gray matter volume and age. Similarly, measures of cardiovascular fitness, as well as objective assessments of physical activity (e.g. actigraphy), should be carried out in future intervention studies, given the limitations of self reported measures. Moreover, exercise interventions should include a long term follow up period in which the same measures obtained at baseline and study

PAGE 48

48 termination are collected. In this way, a physical activity monitoring period can be assessed at foll ow up in an objective fashion (actigraph). 4 1 LIFE P p articipant c haracteristics Physical Activity (N=10) Successful Aging (N=10) M SD M SD t df p M d iff SE d Age 81.11 3.05 81.56 2.77 0.34 18.00 0.74 0.45 1.30 0.04 Edu 14.00 2.83 14.40 2.07 3.61 18.00 0.72 0.40 1.11 0.40 Gender 8 female 7 female 0.27* 1.00 0.61 Kcal 2504.79 1426.11 1587.23 905.83 1.72 18.00 0.10 917.56 534.26 0.19 DS A 0.81 0.15 0.66 0.27 1.47 14.18 0.16 0.15 0.10 0.21 DS RT 1836.79 356.84 2167.14 507.12 1.68 18.00 0.11 330.36 196.09 0.19 Notes: M = mean; SD = standard deviation; df = degrees of freedom; M diff = mean difference; SE = standard error of the difference; d d ; Edu = education; Kcal = to tal amount of calories expended in all exercise related activities (CHAMPS); DS A = digit symbol substitution test accuracy; DS RT = digit symbol substitution test reaction time. = Pearson Chi Square value; df: equal variances not assumed

PAGE 49

49 CHAPTER 5 GENERAL DISCUSSION Age related cognitive decline occurs as the aging brain experiences decrements in gray and white matter volumes, which affect networks needed to perform higher order cognitive functions. It has been shown in the literature that exerc ise and increased levels of physical activity can mitigate brain aging and thus help protect against cognitive decline in late life. The current study investigated the differential effect of self reported physical activity and fitness level on gray matter volume using VBM, while concurrently measuring the direct association between gray matter volume and executive function in a cross sectional sample of older adults. In an attempt to verify our findings, we analyzed structural brain data from a sub sample o f an independent exercise intervention study (LIFE P) which also collected performance on a measure of executive function in older adults previously assigned to a physical activity (PA) or a successful aging health education intervention (SA). Findings from the cross sectional study suggest that fitness level moderat ed the relationship between age and gray matter volume Fitness level displayed a protective role on gray matter volume as a function of age bilaterally in the cerebellum and fusiform/inferio r lateral occipital cortex/lingual gyrus, and left hippocampus and parahippocampal gyrus. These regions have been shown to decline as a function of age, and we were able to find that fitness level mitigated this effect in our cross sectional sample. Execut ive function did not predict the amount of gray matter volume in any area of the brain, nor did gray matter volume predict executive function performance above and beyond the contribution of age. Lack of power within the LIFE P follow up study prevented us from finding significant differences between the PA and

PAGE 50

50 SA groups at the whole brain level and behaviorally. This problem was possibly exacerbated by the fact that the measure used to assign participants to groups was based on self reported physical activ ity and adherence to an intervention that took place two years prior to the current study. In sum, the current study shows that, consistent with the literature, fitness level provided a protective role against brain aging by sparing gray matter tissue lo ss in those with higher fitness level in a cross sectional sample. The same was not true for data obtained from an exercise intervention study; however the lack of findings is most likely due to our small sample size and the self report nature of the measu re used to select the groups. Future exercise and aging studies should be interventional in nature to capture intra and inter individual change in brain structure, function, and cognitive performance. Moreover, future investigations should study the change s that occur in the relationship between gray and white matter volumes as fitness level increases. This would help us to better understand the dynamic changes that occur as the aging brain becomes more plastic following exercise. Furthermore, fitness (VO2 m ax ), objective physical activity (i.e., actigraphy), and cognition should be properly assessed and standardized in the literature to facilitate comparisons between studies and thus reduce methodological variability. Future exercise intervention studies sho uld rely on objective measures of fitness level (i.e., VO2 max or estimates thereof) rather than on self reported measures of physical activity.

PAGE 51

51 APPENDIX 6 1 Median e ffect s izes per region of interest (c ross sectional s tudy)

PAGE 52

52 6 2 Median e ffect s izes per r egion of i nterest (LIFE P)

PAGE 53

53 REFERENCE LIST Andersson, J. L. R., Jenkinson, M., & Smith, S. (2007a). Non linear optimisation. FMRIB technical report TR07JA1 from www.fmrib.ox.ac.uk/analysis/techrep Andersson, J. L. R., Jenkinson, M., & Smi th, S. (2007b). Non linear registration, aka Spatial normalisation. FMRIB technical report TR07JA2 from www.fmrib.ox.ac.uk/analysis/techrep Ashburner, J., & Friston, K. J. (2000). Voxel Based Morphometry : The Methods. NeuroImage, 11 (6), 805 821. Barnes, D E., Yaffe, K., Satariano, W. A., & Tager, I. B. (2003). A Longitudinal Study of Cardiorespiratory Fitness and Cognitive Function in Healthy Older Adults. Journal of the American Geriatrics Society, 51 (4), 459 465. Berchtold, N. C., Chinn, G., Chou, M., Kesslak, J. P., & Cotman, C. W. (2005). Exercise primes a molecular memory for brain derived neurotrophic factor protein induction in the rat hippocampus. Neuroscience, 133 (3), 853 861. Burdette, J. H., Laurienti, P. J., Espeland, M. A., Morgan, A., Telesford, Q., Vechlekar, C. D., et al. (2010). Using Network Science to Evaluate Exercise Associated Brain Changes in Older Adults. Frontiers in Aging Neuroscience, 2 1 10. Colcombe, S., Erickson, K., Raz, N., Webb, A. G. Cohen, N. J., McAuley, E., et al. (2003). Aerobic Fitness Reduces Brain Tissue Loss in Aging Humans. J Gerontol A Biol Sci Med Sci, 58 (2), M176 180. Colcombe, S., Erickson, K., Scalf, P., Kim, J., Prakash, R., McAuley, E., et al. (2006). Aerobic Exercise Training Increases Brain Volume in Aging Humans. J Gerontol A Biol Sci Med Sci, 61 (11), 1166 1170. Colcombe, S., Kramer, A., McAuley, E., Erickson, K., & Scalf, P. (2004). Neurocognitive aging and cardiovascular fitness. Journal of Molecular Neuroscience, 24 (1), 9 14. Colcombe, S., & Kramer, A. F. (2003). Fitness effects on the cognitive function of older adults: A meta analytic study. Psychological Science, 14 (2), 125 130. Colcombe, S., Kramer, A. F., Erickson, K. I., Scalf, P., McAuley, E., Cohen, N. J., et al. (2004). Cardiovascular fitness, cortical plasticity, and aging. Proceedings of the National Academy of Sciences, 101 (9), 3316 3321. Cooper, K. H. (1968). A means of assessing maximal oxygen intake. Correlation between field and treadmill testing. J AMA, 203 (3), 201 204. Dempster, F. N. (1992). The rise and fall of the inhibitory mechanism: Toward a unified theory of cognitive development and aging. Developmental Review, 12 45 75.

PAGE 54

54 Erickson, K. I., Prakash, R., Voss, M. W., Chaddock, L., Hu, L., Morris, K. S., et al. (2009). Aerobic Fitness is Associated With Hippocampal Volume in Elderly Humans. Hippocampus, 19 1030 1039. Erickson, K. I., Raji, C. A., Lopez, O. L., Becker, J. T., R osano, C., Newman, A. B., et al. (2010). Physical activity predicts gray matter volume in late adulthood. Neurology, 75 (16), 1415 1422. Erickson, K. I., Voss, M. W., Prakash, R. S., Basak, C., Szabo, A., Chaddock, L., et al. (2011). Exercise training incre ases size of hippocampus and improves memory. Proceedings of the National Academy of Sciences, 108 (7), 3017 3022. Flel, A., Ruscheweyh, R., Krger, K., Willemer, C., Winter, B., Vlker, K., et al. (2010). Physical activity and memory functions: Are neurot rophins and cerebral gray matter volume the missing link? NeuroImage, 49 (3), 2756 2763. method for grading the cognitive status of patients for the clinician. Journal of Psychiatry Research, 12 (3), 189 198. Godin, G., Jobin, J., & Bouillon, J. (1986). Assessment of leisure time exercise behavior by self report: A concurrent validity study Canadian Journal of Public Health, 77 (5), 359 362. Godin, G., & Shephard, R. J. (1985). A simple method to assess exercise behavior in the community Canadian Journal of Applied Sport Sciences, 10 141 146. Good, C. D., Johnsrude, I. S., Ashburner, J., Henson, R. N. A., Friston, K. J., & Frackowiak, R. S. J. (2001). A Voxel Based Morp hometric Study of Ageing in 465 Normal Adult Human Brains. NeuroImage, 14 (1), 21 36. Gordon, B. A., Rykhlevskaia, E. I., Brumback, C. R., Lee, Y., Elavsky, S., Konopack, J. F., et al. (2008). Neuroanatomical correlates of aging, cardiopulmonary fitness lev el, and education. Psychophysiology, 45 (5), 825 838. Hillman, C. H., Erickson, K. I., & Kramer, A. F. (2008). Be smart, exercise your heart: exercise effects on brain and cognition. Nat Rev Neurosci, 9 (1), 58 65. Hofer, S. M., & Alwin, D. F. (Eds.). (2008) Handbook of Cognitive Aging: Interdisciplinary Perspectives Thousand Oaks: California: Sage Publications, Inc. Honea, R. A., Thomas, G. P., Harsha, A., Anderson, H. S., Donnelly, J. E., Brooks, W. M., et al. (2009). Cardiorespiratory Fitness and Preserv ed Medial Temporal Lobe Volume in Alzheimer Disease. Alzheimer Disease & Associated Disorders, 23 (3), 188 197.

PAGE 55

55 Jacobs, D. R., Ainsworth, B. E., Hartman, T. J., & Leon, A. S. (1993). A simultaneous evaluation of 10 commonly used physical activity questionnaires. Medicine & Science in Sports & Exercise, 25 81 91. Jarvis, B. G. (2004a). MediaLab (Version v2004) New York, NY: Empirisoft Coorporation. Jarvis, B. G. ( 2004b). DirectRT (Version v2004) New York, NY: Empirisoft Coorporation. Kempler, D., & Zelinski, E. M. (1994). Language in dementia and normal aging. In F. A. Dementia and Normal Aging (pp. 331 365). Cambridge, UK: Cambridge University Press. Kramer, A. F., & Erickson, K. I. (2007). Capitalizing on cortical plasticity: influence of physical activity on cognition and brain function. Trends in Cognitive Sciences, 11 (8), 342 348. Kramer, A. F., Erickson, K. I., & Co lcombe, S. J. (2006). Exercise, cognition, and the aging brain. J Appl Physiol, 101 (4), 1237 1242. Mayr, U., & Kliegl, R. (2000). Complex semantic processing in old age: Does it stay or does it go? Psychology and Aging, 15 29 43. McAuley, E., Kramer, A. F ., & Colcombe, S. J. (2004). Cardiovascular fitness and neurocognitive function in older Adults: a brief review. Brain, Behavior, and Immunity, 18 (3), 214 220. McAuley, E., Shaffer, S. M., & Rudolph, D. (1995). Affective responses to acute exercise in elde rly impaired males: the moderating effects of self efficacy and age. International Journal of Aging and Human Development, 41 13 27. Obler, L. K., & Albert, M. L. (Eds.). (1980). Narrative discourse style in the elderly Lexington, MA: Heath. Pahor, M., B lair, S. M., Espeland, M. A., Fielding, R., Gill, T. M., Guralnik, J. M., et al. (2006). Effects of a Physical Activity Intervention on Measures of Physical Performance: Results of the Lifestyle Interventions and Independence for Elders Pilot (LIFE P) Stud y. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 61 (11), 1157 1165. PASW/SPSS, I. (2009). SPSS for Windows Chicago:IL. Pedhazur, E. J. (1997). Multiple Regression in Behavioral Research: Explanation and Prediction (3rd ed .). Orlando: FL: Harcourt Brace College Publishers.

PAGE 56

56 Pereira, A. C., Huddleston, D. E., Brickman, A. M., Sosunov, A. A., Hen, R., McKhann, G. M., et al. (2007). An in vivo correlate of exercise induced neurogenesis in the adult dentate gyrus. Proceedings of the National Academy of Sciences, 104 (13), 5638 5643. Raz, N. (2000). Aging of the Brain and its Impact on Cognitive Performance: Integration of Structural and Functional Findings (Vol. 2). Mahweh, NJ: Lawrence Erlbaum Associates. Raz, N., & Rodrigue, K. M. (2006). Differential aging of the brain: Patterns, cognitive correlates and modifiers. Neuroscience & Biobehavioral Reviews, 30 (6), 730 748. Resnick, S. M., Pham, D. L., Kraut, M. A., Zonderman, A. B., & Davatzikos, C. (2003). Longitudinal Magnetic Reso nance Imaging Studies of Older Adults: A Shrinking Brain. J. Neurosci., 23 (8), 3295 3301. Rosano, C., Venkatraman, V. K., Guralnik, J., Newman, A. B., Glynn, N. W., Launer, L., et al. (2010). Psychomotor Speed and Functional Brain MRI 2 Years After Completing a Physical Activity Treatment. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 65A (6), 639 647. Rueckert, D., Sonoda, L. I., Hayes, C., Hill, D. L. G., Leach, M. O., & Hawkes, D. J. (1999). Non rigid registration using free form deformations: Application to breast MR images. IEEE Transactions on Medical Imaging, 18 (8), 712 721. Ruscheweyh, R., Willemer, C., Krger, K., Duning, T., Warnecke, T., Sommer J., et al. (in press). Physical activity and memory functions: An interventional study. Neurobiology of Aging In Press, Corrected Proof. Salthouse, T. A. (2001). What and When of Cognitive Aging. Current Directions in Psychological Science, 13 (4), 140 1 44. Smith, S. M. (2002). Fast robust automated brain extraction. Human Brain Mapping, 17 (3), 143 155. Smith, S. M., Jenkinson, M., Woolrich, M. W., Beckmann, C. F., Behrens, T. E. J., Johansen Berg, H., et al. (2004). Advances in functional and structural MR image analysis and implementation as FSL. NeuroImage, 23 208 219. Smith, S. M. & Nichols, T. E. (2009). Threshold free cluster enhancement: Addressing problems of smoothing, threshold dependence and localisation in cluster inference Neuroimage, 44 (1), 83 98. Spirduso, W. W. (1975). Reaction and movement time as a function of age and physical activity level. Journal of Gerontology, 30 18 23. Spirduso, W. W., & Clifford, P. (1978). Replication of age and physical activity effects on reaction and moveme nt time. Journal of Gerontology, 1 26 30.

PAGE 57

57 Statistics, F. I. F. o. A. R. (2006). Older Americans Update 2006: Key Indicators of Well Being. In F. I. F. o. A. R. Statistics (Ed.). Washington, DC: U.S.: Government Printing Office. May 2006. Stewart, A. L., V erboncoeur, C. J., & McLellan, B. Y. (2001). Physical activity outcomes of CHAMPS II: a physical activity promotion program for older adults J Gerontol A Biol Sci Med Sci., 56 (A), 465 470. Swain, R. A., Harris, A. B., Wiener, E. C., Dutka, M. V., Morris, H. D., Theien, B. E., et al. (2003). Prolonged exercise induces angiogenesis and increases cerebral blood volume in primary motor cortex of the rat. Neuroscience, 117 (4), 1037 1046. Tisserand, D. J., van Boxtel, M. P. J., Pruessner, J. C., Hofman, P., Evan s, A. C., & Jolles, J. (2004). A Voxel based Morphometric Study to Determine Individual Differences in Gray Matter Density Associated with Age and Cognitive Change Over Time. Cereb ral Cortex, 14 (9), 966 973. Vaynman, S., & Gomez Pinilla, F. (2005). License to Run: Exercise Impacts Functional Plasticity in the Intact and Injured Central Nervous System by Using Neurotrophins. Neurorehabil Neural Repair, 19 (4), 283 295. Vaynman, S., Ying, Z., & Gomez Pinilla, F. (2004). Hippocampal BDNF mediates the efficacy of exercise on synaptic plasticity and cognition. Eu ropean Journal of Neuroscience, 20 (10), 2580 2590. Voss, M. W., Erickson, K. I., Prakash, R. S., Chaddock, L., Malkowski, E., Alves, H., e t al. (2010). Functional connectivity: A source of variance in the association between cardiorespiratory fitness and cognition? Neuropsychologia, 48 (5), 1394 1406. West, R. (1996). An Application of Prefrontal Cortex Theory to Cognitive Aging. American Psy chological Association, 120 272 292. Wilson, R. S., Bennett, D. A., Bienias, J. L., Aggarwal, N. T., Mendes de Leon, C. F., Morris, M. C., Schneider, J. A., Evans, D. A. (2002). Cognitive Activity and Incident AD in a Population Based Sample of Older Adul ts. Neurology, 59 1910 1914. Zhang, Y., Brady, M., & Smith, S. M. (2001). Segmentation of brain MR images through a hidden Markov random field model and the expectation maximization algorithm. IEEE Trans. on Medical Imaging, 20 (1), 45 57.

PAGE 58

58 BIOGRAP HICAL SKETCH Zvinka Zoe Zlatar received her Bachelor of Arts degree in psychology from the University of Nevada, Las Vegas in 2005 and her Master of Science degree in psychology from the University of Florida in the spring of 2009. Zvinka is currently a fo urth year neuropsychology graduate student in the University of Florida's Clinical and Health Psychology program. She works in the Brain Imaging Rehabilitation and Cognition lab under the supervision of Bruce Crosson, Ph.D. and has been funded by the Natio nal Institute on Aging through both a supplemental grant awarded through the Claude D. Pepper Older Americans Independence Center, and the University of studying neurocogniti ve aging in healthy older adults using both functional and structural magnetic resonance imaging (MRI). Zvinka is also interested in investigating life style factors that can potentially mitigate age related cognitive decline, such as exercise and chang es in physical activity levels.