Long-term mild-intensity exercise regimen preserves prefrontal cortical volume against aging

RESEARCH ARTICLE

Long-term mild-intensity exercise regimen preserves prefrontal cortical volume against aging Masashi Tamura1,2, Kiyotaka Nemoto3, Atsushi Kawaguchi4, Morimasa Kato5, Tetsuaki Arai3, Tatsuyuki Kakuma6, Katsuyoshi Mizukami7, Hiroshi Matsuda8, Hideaki Soya9 and Takashi Asada3 1

Department of Neuropsychiatry, Graduate School of Comprehensive Human Sciences, University of Tsukuba, Tsukuba, Ibaraki Japan Ibaraki Prefectural Medical Center of Psychiatry, Kasama, Ibaraki Japan 3 Department of Psychiatry, Division of Clinical Medicine, Faculty of Medicine, University of Tsukuba, Tsukuba, Ibaraki Japan 4 Department of Biomedical Statistics and Bioinformatics, Kyoto University Graduate School of Medicine, Kyoto, Japan 5 Department of Health and Nutrition, Yonezawa Women’s Junior College of Yamagata Prefecture, Yonezawa, Yamagata Japan 6 Biostatistics Center, Kurume University, Kurume, Fukuoka Japan 7 Graduate School of Comprehensive Human Sciences, Faculty of Health and Sports Sciences, University of Tsukuba, Tokyo, Japan 8 Integrative Brain Image Center, National Center of Neurology and Psychiatry, Tokyo, Japan 9 Laboratory of Exercise Biochemistry and Neuroendocrinology, Faculty of Health and Sports Sciences, University of Tsukuba, Tsukuba, Ibaraki Japan Correspondence to: K. Nemoto and H. Soya, E-mail: [email protected]; [email protected] 2

It has been suggested that exercise improves cognitive function and increases cerebral volume even in older people. However, the relation between cognitive function and brain volume is unclear. We evaluated the longitudinal change of cognitive function and gray matter volume due to mild-intensity exercise over 2 years, and the residual effects 6 months post-exercise. Methods: Subjects were 110 healthy older individuals over 65 years old in Tone town, Ibaraki prefecture. Seventy-five participants were voluntarily enrolled in the exercise group. A mild-intensity calisthenics regimen, which consisted of home-based and club-based programs for as long as 2 years, was employed as the intervention for the exercise group. Results: The exercise group showed significant improvement in attentional shift over the course of the observation period including a 6-month follow-up. Neuroimaging analysis revealed the significant preservation of bilateral prefrontal volume in the exercise group with small-volume corrections, although this effect faded after intervention. Furthermore, the longitudinal changes in attentional shift and memory were positively correlated with the prefrontal volumetric changes. Conclusion: Our results suggest that mild-intensity exercise could prevent prefrontal volume reduction due to aging and impede cognitive decline. Copyright # 2014 John Wiley & Sons, Ltd. Objectives:

Key words: aging; brain; exercise; cognition; voxel-based morphometry; longitudinal study History: Received 6 March 2014; Accepted 5 August 2014; Published online in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/gps.4205

Introduction Accumulated evidence has suggested that physical activity (PA) can improve cognitive functions, such as executive functions, attention, and processing speed, even in older people (Angevaren et al., 2008; van Uffelen et al., 2008; Smith et al., 2010). Some studies have investigated the effect of PA on the aging brain. For example, Colcombe et al. (2006) found that aerobic fitness maintained for 6 months Copyright # 2014 John Wiley & Sons, Ltd.

increased the volume primarily located in prefrontal and temporal cortices. Erickson et al. (2011) also found that 1 year of aerobic exercise increased the bilateral anterior hippocampal volume. Ruscheweyh et al. (2011) reported that 6 months of aerobic exercise resulted in a positive association between the increase in total PA and increases in prefrontal and cingulate cortical volumes. These studies suggest that exercise can influence the plasticity of the aging brain. Int J Geriatr Psychiatry 2014

M. Tamura et al.

However, several important questions remain unanswered. First, it is unclear how long-term PA acts in older people and whether its effect remains after intervention. In previous studies, most intervention periods were several months long, and none extended into multi-year periods. Animal studies have suggested that brain volume alteration occurs under the “use it or lose it” hypothesis (Anderson, 2011), but changes in cerebral volume after intervention have not been sufficiently investigated in humans. Second, the association between the intensity of PA and cerebral plasticity as well as cognitive change has not been investigated sufficiently. Most previous studies have employed aerobic exercise for the intervention. Ruscheweyh et al. (2011) observed a differential effect of low-intensity and moderate-intensity exercise, suggesting that the increase of total PA mattered more than the intensity. Particularly in older people, maintaining a regular exercise regimen is often not feasible because of a lack of motivation, a lack of opportunities, or cardiovascular and musculoskeletal deficiencies (Irwin et al., 2004). Therefore, mild-intensity exercise might be preferable in terms of both adoption and maintenance. Timinkul et al. (2008) investigated the influence of exercise intensity on the prefrontal cerebral blood volume in healthy young men and found that oxygenation of the prefrontal cortex increased during graded cycling even at a mild intensity. Although these studies imply the potential of mildintensity exercise, more findings are needed to explore this. Third, the relationship among PA, cognitive function, and regional brain volume is not fully understood. Previous neuroimaging studies that evaluated cognitive change due to PA have focused on specific domains of cognition such as spatial memory (Erickson et al., 2011) or episodic memory (Ruscheweyh et al., 2011). Addressing these questions is essential in the context of the prevention of cognitive decline and brain atrophy with aging. Therefore, we designed a 2-year mild-intensity exercise intervention followed by 6 months of observation for healthy older subjects and investigated the association between PA, cognitive function, and regional cerebral volume. We hypothesized that although mild-intensity PA would have positive effects on cognitive function and regional brain volume, its effect could be transient, and volume would decrease after intervention.

investigate cognitive function and the prevention of dementia in a community (Miyamoto et al., 2009). This project was approved by the ethical committee of Tsukuba University. The candidates for this project were 2698 inhabitants aged 65 years and older living in Tone town as of 1 December 2001, and 1888 of these participated in the study. At baseline, we performed a screening survey to assess the cognitive function of participants (Miyamoto et al., 2009; Yasuno et al., 2012). During the screening survey, all participants underwent a structured interview, including questions about age, sex, education, and physical and psychiatric conditions. Furthermore, in order to evaluate cognitive function, we used the 5-cognitive function test (5-Cog), which measures five domains of cognition: attentional shift, memory, visuospatial function, verbal fluency, and reasoning. The 5-Cog is described in the following. We explained the present study to 1888 participants and obtained written informed consent in accord with the ethical guidelines of the local ethical committee. Participants were recruited between 1 April 2002 and 30 September 2002. Of these, 347 agreed to undergo magnetic resonance imaging (MRI) brain scans. All MRI scans were performed using the same scanner at the JA Toride Medical Center. Participants also underwent a Mini-mental state examination (MMSE) on the same date as their MRI scan. Of the 347 subjects, 138 were excluded as they had a history of psychiatric disease, stroke, or grade 3 white matter hyperintensity as rated using the Fazekas scale (Fazekas et al., 1993) on axial T2 images at baseline. We considered those who met both of the following conditions as cognitively normal: (1) all domain scores of the 5-Cog at baseline were above the average 1 standard deviation calculated from all 1888 participants of the Tone Project; and (2) all MMSE scores (from three time points within the observation period) were 24 or above. Fifty-three subjects were excluded because they did not meet the said conditions. As a result, 156 subjects remained for the present study. They were divided into two groups on a voluntary basis: exercise intervention group (n = 102) and control group (n = 54). Among these candidates, 27 in the exercise group and 19 in the control group were excluded from analyses as they did not undergo all three MRIs. As a result, 110 subjects (exercise = 75 and control = 35) remained for analyses (Figure 1).

Methods Subjects

Exercise intervention

Participants were recruited from the “Tone Project” in Tone town, Ibaraki, Japan, which we started in 2001 to

The exercise intervention was held over 2 years, from 1 April 2003 to 31 March 2005. In order for the

Copyright # 2014 John Wiley & Sons, Ltd.

Int J Geriatr Psychiatry 2014

Benefits of ongoing mild exercise to aging brain

Figure 1 Flow chart of subject recruitment.

participants to maintain their exercise, we developed a mild and enjoyable exercise regimen. The exercise course consisted of home-based and club-based programs. Home-based exercise was the daily practice of mild-intensity calisthenics, called “furi-furi-guppa” in Japanese: with the knees bent, one sways the hips from side to side with the arms open and hands clapping in tandem. The intensity of this calisthenics was expected to be 4.5 metabolic equivalents. We asked the participants to do calisthenics with their favorite songs for 10 min at a time, three times a day. Each participant recorded the time devoted exclusively to the calisthenics. The club-based exercise program was held monthly at the community center. This program was 1 h of exercise consisting of four sections: warm-up and stretching (10 min), group exercise using rubber balls (30 min), calisthenics (10 min), and cool down (10 min). After the exercise, we interviewed the participants and reviewed their records for the home-based program. After the intervention (March 2006), we performed a follow-up survey to examine whether subjects in the exercise group had continued the calisthenics.

reasoning. We evaluated attentional shift using a Japanese version of a set-dependent activity (Sohlberg and Matter, 1986). In this test, three words, “top,” “middle,” or “bottom,” are randomly placed in three rows, top, middle, and bottom, so that the meaning of each word does not necessarily correspond to its position. The participants were required to choose the words that were placed in the correct rows. Memory was assessed using a Category Cued Recall test (Grober et al., 1988). A Clock Drawing test (Freedman et al., 1994) was used to assess visuospatial function. We examined language ability using a category fluency test (Solomon and Pendlebury, 1998). To assess reasoning ability, we employed the similarity subset of the Wechsler Adult Intelligence Scale-Revised (Wechsler, 1981). We calculated a composite score from raw 5-Cog scores. The formula for the composite score was described elsewhere (Yasuno et al., 2012). Because 5-Cog and MRI were not performed on the same days (mean time gap was 67.1 ± 44.7 days for the exercise group and 61.2 ± 50.1 days for the control group), we estimated the 5-Cog scores for the day subjects underwent brain MRIs using ordinary least squares. We used these estimated scores for the analyses.

Cognitive assessment Magnetic resonance imaging acquisition

All participants underwent 5-Cog annually. 5-Cog measures five domains of cognitive function: attentional shift, memory, visuospatial ability, verbal fluency, and Copyright # 2014 John Wiley & Sons, Ltd.

All MRI scans were performed with a 1.5-T MRI scanner (Symphony, Siemens, Germany). A threeInt J Geriatr Psychiatry 2014

M. Tamura et al.

dimensional volumetric acquisition of a T1-weighted gradient echo sequence produced a gapless series of thin sagittal sections using an MPRAGE sequence (echo time/repetition time, 3.93/2800 ms; flip angle, 12°; acquisition matrix, 512 × 512; field of view, 28.0 cm; slice thickness, 1.20 mm). All subjects underwent three MRIs. The first scans (pre-intervention: pre) were taken as a baseline at the onset of the program, the second scans (intermediate: int) were taken approximately 1 year into the intervention (average interval between pre and int: 403.3 ± 48.4 days), and the third scans (post-intervention: post) were taken approximately 6 months after the intervention ended (average interval between pre and post: 927.3 ± 69.1 days). Two-sample T-tests showed no significant differences of MRI scan intervals between the groups. Magnetic resonance imaging analysis

Before preprocessing of the voxel-based morphometry (VBM), we employed N3 1.10 (Sled et al., 1998) to correct the intensity inhomogeneity of all MRI scans. The N3 correction process is described elsewhere (Nemoto et al., 2011). After inhomogeneity correction, we used SPM8 (Wellcome Institute of Neurology, University College London, UK), running on MATLAB R2010a (MathWorks, Natick, MA, USA) and VBM8 toolbox (http://dbm.neuro.uni-jena.de/vbm/), for the preprocessing of the images. As magnetic resonance images were longitudinal data, we used the longitudinal preprocessing script implemented in the VBM8 toolbox. In brief, after an initial realignment of pre, int, and post images, the mean of the realigned images was calculated and used as a reference image in subsequent realignments. The realigned images were then corrected for signal inhomogeneities with regard to the reference mean image. Spatial normalization parameters with a fast diffeomorphic image registration algorithm (Ashburner, 2007) were estimated using the segmentations (gray matter (GM) and white matter images) of the mean image. These normalization parameters were applied to the segmentations of the bias-corrected images. The resulting normalized segmentations were again realigned. Finally, GM images were smoothed using an 8-mm full-width half maximum of isotropic Gaussian kernel. Statistical analysis

Statistical analyses were performed with SPSS Statistics 19 (IBM Japan, Tokyo) and SPM8, which implemented Copyright # 2014 John Wiley & Sons, Ltd.

a general linear model. First, we performed propensity score analyses to balance measurable confounders between the groups. Propensity score takes confounding variables by indication into account (Rosenbaum and Rubin, 1983). Propensity is defined as an individual’s probability of being positively affected by the intervention of interest given the complete set of all information about that individual. The propensity score provides a single metric that summarizes all the information from explanatory variables (Nicholas and Gulliford, 2008). In order to calculate the propensity score, a logistic regression using SPSS was used on the confounding covariates, including sex, years of education, age, MMSE score, and 5-Cog composite score at baseline. This propensity score was used as a covariate to adjust the group differences in the following statistical analyses. Austin (2011a) provides a detailed guide for analyses with propensity score. For longitudinal VBM with a hypothesis based on previous reports that exercise affects prefrontal cortices (PFCs) and hippocampi, we investigated regional changes in PFCs and hippocampi using small-volume correction with WFU PICKATLAS software (Maldjian et al., 2003) as well as whole-brain analysis. Significance levels were set at p < 0.05 (family-wise error (FWE) corrected). We obtained both MNI and Talairach coordinates to detect the anatomical region of the clusters. We used a transform from Matthew Brett (http://imaging.mrc-cbu.cam.ac.uk/ imaging/MniTalairach) to convert MNI coordinates to Talairach coordinates. Talairach Client 2.4.3 (Lancaster et al., 2000) was used to identify the anatomical regions corresponding to Talairach coordinates. Longitudinal changes of cognition and regional gray matter volume. In order to evaluate the effect of exer-

cise on cognition, we performed a two-way repeatedmeasures analysis of variance (ANOVA) using SPSS (IBM Corporation, Armonk, NY, USA). For this analysis, group (between subject) and time (within subject) were entered as independent variables, and the composite score and each domain score for 5-Cog were used as dependent variables. We treated the propensity score as a covariate. The Tukey honestly significant difference test was performed post hoc. For longitudinal VBM, we used the full-factorial design and weighted contrast vectors for estimated parameters. To model our hypothesis of exercise effect leading to increased brain volume between pre and int followed by a decrease between int and post, we set a contrast vector of [ 0.5 1 0.5] in the pre, int, and post columns for the exercise group. To observe the interaction between the two groups, a contrast vector Int J Geriatr Psychiatry 2014

Benefits of ongoing mild exercise to aging brain

of [0.5 1 0.5] was set on parameters of the pre, int, and post columns for the control group. McDonald et al. (2010) described in detail how to model weighted contrast vectors. The propensity score was treated as a covariate. Correlation between cognition and regional gray matter volume. We used the multiple regression model to

separately for the two groups, treating age at baseline and years of education as covariates. Results Demographics

analyze the correlation between regional GM volume change and cognitive change, using subtraction images and each domain score of 5-Cog between pre and int. The subtraction images were generated using FSL 4.1.9 (FMRIB Software Library, The University of Oxford; Woolrich et al., 2009). Analyses were performed

Subject demographics are shown in Table 1. Although there were no significant differences in age or sex, significant differences were found in years of education, MMSE, and 5-Cog scores between groups at baseline. Of the 75 subjects in the exercise group, 15 (20%) had continued the calisthenics at home at the point of the follow-up interview in 2006.

Table 1 Demographic information for subjects at baseline

Longitudinal changes of cognition and regional gray matter volume

Age Years of education Gender Mini-mental state examination Composite score of 5-Cog

Exercise (n = 75)

Control (n = 35)

Mean ± SD

Mean ± SD

p-value

72.0 ± 4.1 11.6 ± 2.7

72.0 ± 4.7 10.1 ± 2.5

0.974 0.004

46 female (61%) 28.9 ± 1.3

14 female (40%) 28.0 ± 2.0

0.059

59.1 ± 11.3

49.3 ± 14.1

0.025 0.05). However, when we focused on the prefrontal regions with smallvolume correction, we found a significant interaction in the bilateral middle frontal gyri (Brodmann area (BA) 10). Their volumes were preserved between pre (baseline) and int (1 year into the intervention) and reduced between int and post (6 months after the intervention ended) in the exercise group, whereas they decreased consistently over time in the control group (Figure 3 and Table 2). Nevertheless, there was no significant difference in these regions between the groups after the intervention.

Correlation between cognition and regional gray matter volume

For the exercise group, we found that changes in attentional shift scores were positively correlated with the bilateral superior frontal gyri (BA 10) and right middle frontal gyrus (BA 46; Figure 4a and Table 2). Changes in memory scores were also positively correlated with the left superior (BA 10) and inferior frontal gyri (BA 9; Figure 4b and Table 2). These regions partially overlapped with the volume-preserved areas during the intervention (Figure 3). Changes in other cognitive domains showed no significant association with any regional

GM volume including hippocampi. On the other hand, no significant correlation was observed in the control group. Table 2 Gray matter regions detected in each analysis

Region

BA

MNI coordinates

k

Gray matter regions with group-by-time interaction Rt. middle 10 30 54 8 53 frontal gyrus Lt. middle 10 29 50 5 51 frontal gyrus

t statistics

3.37 3.57

Clusters positively correlating to changes in attentional shift scores in the exercise group Rt. middle 46 50 47 6 74 3.76 frontal gyrus (Cluster 1) Rt. superior 10 26 60 14 180 3.98 frontal gyrus (Cluster 2) 10 32 56 11 116 3.71 Lt. superior frontal gyrus (Cluster 3) Clusters positively correlating to changes in memory scores in the exercise group Lt. superior 10 23 60 18 385 4.74 frontal gyrus (Cluster 4) Lt. inferior frontal 9 59 8 30 207 3.95 gyrus (Cluster 5) Significant clusters at the voxel level family-wise error-corrected p < 0.05 with small-volume correction. BA, Brodmann area; Rt, right; Lt, left.

Figure 3 Regional gray matter changes of exercise and control groups. The red blobs showed significant group-by-time interaction in the bilateral middle frontal gyri (Brodmann area (BA) 10). In the exercise group, we did not observe significant volume change in these regions between pre (baseline) and int (1 year into the intervention). However, significant reduction was found between int and post (6 months after the intervention). However, volumes of these regions significantly decreased over time in the control group.

Copyright # 2014 John Wiley & Sons, Ltd.

Int J Geriatr Psychiatry 2014

Benefits of ongoing mild exercise to aging brain

Figure 4 Correlation analyses between cognitive change and regional gray matter volume change. (a) The change in attentional shift positively correlated with the right middle frontal gyrus (Brodmann area (BA) 46, Cluster 1) and bilateral superior frontal gyri (BA 10, Clusters 2 and 3) in the exercise group (yellow). Scatter plots show the positive correlation between the volume changes of each cluster and score changes in attentional shift for the exercise group.(b) Changes in memory positively correlated with the left superior frontal gyrus (BA 10, Cluster 4) and left inferior frontal gyrus (BA 9, Cluster 5) in the exercise group (blue). Scatter plots show the positive correlation between the volume changes of each region and score changes in memory for the exercise group.

Discussion We evaluated the long-term effects of a mildintensity exercise on cognitive functions and regional GM volume in older people. We found significant interaction between the exercise and control groups in attentional shift from a cognitive perspective and in bilateral prefrontal GM volume from a neuroimaging perspective. We also found that attentional shift and memory changes were positively correlated with the prefrontal GM volume change during the first year of the Copyright # 2014 John Wiley & Sons, Ltd.

intervention, whereas no significant correlation was found in the control group. The longitudinal effect of exercise on brain regions was transient as we predicted based on the “use it or lose it” hypothesis (Anderson, 2011). On the other hand, attentional shift performance improved over the course of the observation period, including the 6-month follow-up. Our long-term intervention and follow-up observations revealed that exercise seemed to have differential longitudinal effects on cognition and regional brain volume. In Alzheimer’s disease, it is suggested that brain Int J Geriatr Psychiatry 2014

M. Tamura et al.

changes become observable prior to affecting cognitive functions (Swain et al., 2003). If this also applies to the healthy aging brains, the improved attentional shift observed in our study may decline after a longer time span. Nonetheless, our results suggest that mild-intensity exercise might impede regional GM atrophy due to aging, which would result in delaying cognitive decline. As for the correlation between cognition and regional GM volume due to exercise, changes in attentional shift had a positive correlation with volume changes of the PFCs. Attentional shift is proposed as part of the executive function and is functionally localized in the lateral PFC (Mirsky, 1987). Our results coincide with previous findings, and this strengthens the possibility that exercise induces both attentional shift improvement and prefrontal GM volume preservation. In addition, changes in memory were positively associated with prefrontal volume changes, although there was no significant interaction in memory scores between the groups. Activation of the lateral PFC might lead to memory improvement because memory encoding and retrieval are largely mediated by this region (Fletcher and Henson, 2001). Although the hippocampus plays an important role in memory, we did not find any significant correlation. Erickson et al. (2011) reported that moderate-intensity exercise performed over 1 year increased anterior hippocampal volume in the older people. Herting and Nagel (2012) reported that the intensity of exercise had a positive correlation with hippocampal volume in adolescents. Considering these findings together with our results, it seems plausible that hippocampal volume change depends on exercise intensity. It is noteworthy that the average volume of the PFC in the exercise group did not increase but was preserved. The PFC shows the most atrophy in an aging brain (Raz et al., 2004). In fact, the volume decrease of the PFC in the control group is consistent with this. It is suggested that there is not only volume but also functionality decline or change in aging PFC (Bentourkia et al., 2000; Volkow et al., 2000). These results imply that exercise might impede multilevel age-related decline of PFC functions such as glucose metabolism and perfusion, leading to the temporary preservation of PFC volume. Further functional imaging studies using single photon emission computed tomography or positron emission tomography are necessary in order to reveal the mechanism behind exercise effects on brain. Some may be concerned that the mean PFC volume of the exercise group was reduced to the same level as Copyright # 2014 John Wiley & Sons, Ltd.

that of the control group after the intervention. Previous reports indicate that a modest amount of activity is needed for detectable effects on GM volume (Floel et al., 2010; Erickson et al., 2011; Ruscheweyh et al., 2011; Gow et al., 2012), although these studies did not evaluate the long-term effects after intervention. At the same time, it is reported that the intensity of exercise might not be associated with cognitive improvement (Smith et al., 2010; Ruscheweyh et al., 2011). In addition to these reports, our study implies that, after the cessation of exercise, the positive effects of mild-intensity exercise yield to the aging effect relatively soon in terms of morphology, whereas they are maintained for a longer period in terms of cognition. As exercise intensity increases, it becomes more difficult for older people to make it a habit. In light of this, mild and easy-to-continue aerobic exercise might be beneficial for preventing cognitive decline due to aging. There are some limitations in this study. First, we could not conduct a randomized controlled trial. In fact, significant baseline differences were found in the characteristics between the two groups. In order to overcome this problem, we performed a propensity score analysis, which allows one to design and analyze a nonrandomized study so that it mimics some of the particular characteristics of a randomized controlled trial (Austin, 2011b). Therefore, we consider that our results reflect the effects induced by continuous, mildintensity exercise. Second, MRI scanning was not carried out immediately after the intervention, so we could not evaluate cognition and regional GM volume at the end of the intervention. Nevertheless, we were able to assess the effect of exercise for at least 1 year. Lastly, there may have been a practice effect on the memory task, as both groups exhibited significant improvement in memory. However, the control group did not exhibit any significant differences between any time points for the rest of the 5-Cog domains. In conclusion, regular mild-intensity exercise in older people improved attentional shift and impeded prefrontal GM volume reduction due to aging. These results suggest that exercise could delay cognitive decline due to aging. Conflict of interest None declared. Int J Geriatr Psychiatry 2014

Benefits of ongoing mild exercise to aging brain

Key points Two-year mild-intensity exercise intervention improved attentional shift over the course of the observation period, including a 6-month follow-up. Bilateral prefrontal gray matter volume was preserved during the exercise period, although this effect faded after intervention. Attentional shift and memory changes were positively correlated with changes in prefrontal volume during the intervention. Mild-intensity exercise could affect both cognition and regional gray matter volume and delay cognitive decline due to aging.

• • • •

Acknowledgements This research was funded by the Ministry of Health, Labour, and Welfare of Japan (H13-dementia and fracture-003). References Anderson BJ. 2011. Plasticity of gray matter volume: the cellular and synaptic plasticity that underlies volumetric change. Dev Psychobiol 53(5): 456–465. Angevaren M, Aufdemkampe G, Verhaar HJ, et al. 2008. Physical activity and enhanced fitness to improve cognitive function in older people without known cognitive impairment. Cochrane Database Syst Rev (3): CD005381. Ashburner J. 2007. A fast diffeomorphic image registration algorithm. Neuroimage 38(1): 95–113. Austin PC. 2011a. A tutorial and case study in propensity score analysis: an application to estimating the effect of in-hospital smoking cessation counseling on mortality. Multivariate Behav Res 46(1): 119–151. Austin PC. 2011b. An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate Behav Res 46(3): 399–424. Bentourkia M, Bol A, Ivanoiu A, et al. 2000. Comparison of regional cerebral blood flow and glucose metabolism in the normal brain: effect of aging. J Neurol Sci 181(1–2): 19–28. Colcombe SJ, Erickson KI, Scalf PE, et al. 2006. Aerobic exercise training increases brain volume in aging humans. J Gerontol A Biol Sci Med Sci 61(11): 1166–1170. Erickson KI, Voss MW, Prakash RS, et al. 2011. Exercise training increases size of hippocampus and improves memory. Proc Natl Acad Sci U S A 108(7): 3017–3022. Fazekas F, Kleinert R, Offenbacher H, et al. 1993. Pathologic correlates of incidental MRI white matter signal hyperintensities. Neurology 43(9): 1683–1689. Fletcher PC, Henson RN. 2001. Frontal lobes and human memory: insights from functional neuroimaging. Brain 124(Pt 5): 849–881. Floel A, Ruscheweyh R, Kruger K, et al. 2010. Physical activity and memory functions: are neurotrophins and cerebral gray matter volume the missing link? Neuroimage 49(3): 2756–2763.

Copyright # 2014 John Wiley & Sons, Ltd.

Freedman M, Leach L, Kaplan E, et al. 1994. Clock Drawing. A Neuropsychological Analysis. Oxford University Press: New York. Gow AJ, Bastin ME, Munoz Maniega S, et al. 2012. Neuroprotective lifestyles and the aging brain: activity, atrophy, and white matter integrity. Neurology 79(17): 1802–1808. Grober E, Buschke H, Crystal H, et al. 1988. Screening for dementia by memory testing. Neurology 38(6): 900–903. Herting MM, Nagel BJ. 2012. Aerobic fitness relates to learning on a virtual Morris Water Task and hippocampal volume in adolescents. Behav Brain Res 233(2): 517–525. Irwin ML, Tworoger SS, Yasui Y, et al. 2004. Influence of demographic, physiologic, and psychosocial variables on adherence to a yearlong moderate-intensity exercise trial in postmenopausal women. Prev Med 39(6): 1080–1086. Lancaster JL, Woldorff MG, Parsons LM, et al. 2000. Automated Talairach atlas labels for functional brain mapping. Hum Brain Mapp 10(3): 120–131. Maldjian JA, Laurienti PJ, Kraft RA, et al. 2003. An automated method for neuroanatomic and cytoarchitectonic atlas-based interrogation of fMRI data sets. Neuroimage 19(3): 1233–1239. McDonald BC, Conroy SK, Ahles TA, et al. 2010. Gray matter reduction associated with systemic chemotherapy for breast cancer: a prospective MRI study. Breast Cancer Res Treat 123(3): 819–828. Mirsky AF. 1987. Behavioral and psychophysiological markers of disordered attention. Environ Health Perspect 74: 191–199. Miyamoto M, Kodama C, Kinoshita T, et al. 2009. Dementia and mild cognitive impairment among non-responders to a community survey. J Clin Neurosci 16(2): 270–276. Nemoto K, Dan I, Rorden C, et al. 2011. Lin4Neuro: a customized Linux distribution ready for neuroimaging analysis. BMC Med Imaging 11: 3. Nicholas JJ, Gulliford MC. 2008. Commentary: what is propensity score? Br J Gen Pract 58: 687. Raz N, Gunning-Dixon F, Head D, et al. 2004. Aging, sexual dimorphism, and hemispheric asymmetry of the cerebral cortex: replicability of regional differences in volume. Neurobiol Aging 25(3): 377–396. Rosenbaum PR, Rubin DB. 1983. The central role of the propensity score in observational studies for causal effects. Biometrika 70(1): 41–55. Ruscheweyh R, Willemer C, Kruger K, et al. 2011. Physical activity and memory functions: an interventional study. Neurobiol Aging 32(7): 1304–1319. Sled JG, Zijdenbos AP, Evans AC. 1998. A nonparametric method for automatic correction of intensity nonuniformity in MRI data. IEEE Trans Med Imaging 17(1): 87–97. Smith PJ, Blumenthal JA, Hoffman BM, et al. 2010. Aerobic exercise and neurocognitive performance: a meta-analytic review of randomized controlled trials. Psychosom Med 72(3): 239–252. Sohlberg MM, Matter CA. 1986. Attention Process Training (APT). Association for Neuropsychological Research and Development. Solomon PR, Pendlebury WW. 1998. Recognition of Alzheimer’s disease: the 7 minute screen. Fam Med 30(4): 265–271. Swain RA, Harris AB, Wiener EC, et al. 2003. Prolonged exercise induces angiogenesis and increases cerebral blood volume in primary motor cortex of the rat. Neuroscience 117(4): 1037–1046. Timinkul A, Kato M, Omori T, et al. 2008. Enhancing effect of cerebral blood volume by mild exercise in healthy young men: a near-infrared spectroscopy study. Neurosci Res 61(3): 242–248. van Uffelen JG, Chin APMJ, Hopman-Rock M, et al. 2008. The effects of exercise on cognition in older adults with and without cognitive decline: a systematic review. Clin J Sport Med 18(6): 486–500. Volkow ND, Logan J, Fowler JS, et al. 2000. Association between age-related decline in brain dopamine activity and impairment in frontal and cingulate metabolism. Am J Psychiatry 157(1): 75–80. Wechsler D. 1981. WAIS-R: Manual: Wechsler Adult Intelligence Scale-Revised. Harcourt Brace Jovanovich for Psychological Corp: New York. Woolrich MW, Jbabdi S, Patenaude B, et al. 2009. Bayesian analysis of neuroimaging data in FSL. Neuroimage 45(1 Suppl): S173–S186. Yasuno F, Tanimukai S, Sasaki M, et al. 2012. Effect of plasma lipids, hypertension and APOE genotype on cognitive decline. Neurobiol Aging 33(11): 2633–2640.

Int J Geriatr Psychiatry 2014