Contents lists available at SciVerse ScienceDirect
Experimental Gerontology
j ourna l homepage: www.e lsev ie r .com/ locate /expgero
Effect of age on exercise-induced alterations in cognitive executive function:Relationship to cerebral perfusion
Samuel J.E. Lucas a,b,⁎, Philip N. Ainslie c, Carissa J. Murrell a,b, Kate N. Thomas a,Elizabeth A. Franz d, James D. Cotter b
a Department of Physiology, University of Otago, Dunedin, New Zealandb School of Physical Education, University of Otago, Dunedin, New Zealandc School of Health and Exercise Sciences, University of British Columbia, Okanagan Campus, Canadad Department of Psychology and fMRI Otago, University of Otago, Dunedin, New Zealand
Abbreviations: CBF, cerebral blood flow; HHb, deoxrange; MCAv, middle cerebral artery blood flow velocitycopy; O2Hb, oxyhemoglobin; PCO2, partial pressure of Cend tidal CO2; tHb, total hemoglobin; TOI, total corticamaximal aerobic power test.⁎ Corresponding author at: University of Otago Dunedi
Article history:Received 12 October 2011Received in revised form 6 December 2011Accepted 12 December 2011Available online 2 January 2012
Section Editor: Christian Humpel
Keywords:AgingCerebral blood flowNear-infrared spectroscopyExecutive function cognitionExercise
Regular exercise improves the age-related decline in cerebral blood flow (CBF) and is associated with im-proved cognitive function; however, less is known about the direct relationship between CBF and cognitivefunction. We examined the influence of healthy aging on the capability of acute exercise to improve cogni-tion, and whether exercise-induced improvements in cognition are related to CBF and cortical hemodynamics.Middle cerebral artery blood flow velocity (MCAv; Doppler) and cortical hemodynamics (NIRS) were measuredin 13 young (24±5 y) and 9 older (62±3 y) participants at rest and during cycling at 30% and 70% of heart raterange (HRR). Cognitive performance was assessed using a computer-adapted Stroop task (i.e., test of executivefunction cognition) at rest and during exercise. Average response times on the Stroop task were slower forthe older compared to younger group for both simple and difficult tasks (Pb0.01). Independent of age,difficult-task response times improved during exercise (Pb0.01), with the improvement greater at 70% HRR ex-ercise (P=0.04 vs. 30% HRR). Higher MCAv was correlated with faster response times for simple and difficulttasks at rest (R2=0.47 and R2=0.47, respectively), but this relation uncoupled progressively during exercise.Exercise-induced increases in MCAv were similar and unaltered during cognitive tasks for both age groups. Incontrast, prefrontal cortical hemodynamic NIRS measures [oxyhemoglobin (O2Hb) and total hemoglobin(tHb)]were differentially affected by exercise intensity, age and cognitive task; e.g., therewere smaller increasesin [O2Hb] and [tHb] in the older group between exercise intensities (Pb0.05). These data indicate that: 1) Re-gardless of age, cognitive (executive) function is improved while exercising; 2) while MCAv is strongly relatedto cognition at rest, this relation becomes uncoupled during exercise, and 3) there is dissociation between globalCBF and regional cortical oxygenation and NIRS blood volume markers during exercise and engagement of pre-frontal cognition.
Aging is reflected in progressive declines in cognitive function(Park et al., 2001) and cerebral blood flow (CBF) (Ainslie et al.,2008; Bertsch et al., 2009; Buijs et al., 1998). Exercise participationand higher fitness levels have consistently emerged as key mediatorsof improved cognitive function (Abbott et al., 2004; Colcombe andKramer, 2003; Hill et al., 1993; Kramer et al., 1999; Voss et al.,2011; Weuve et al., 2004) and CBF (Ainslie et al., 2008; Rhyu et al.,
2010). While the natural age-related declines in resting CBF havebeen associated with poorer cognitive function (Bertsch et al., 2009;Heo et al., 2010), few studies have investigated whether acute alter-ations in cerebral perfusion directly affect cognitive function. In sup-port of this possibility, Marshall et al. (2001) reported that thetransient occlusion of CBF in patients with cerebrovascular diseaseimpaired cognition (attention) until the occlusion was removed. Todate, however, no studies have investigated whether an acute in-crease in CBF is directly related to improved cognitive performancein healthy younger and older individuals.
We (Lucas et al., 2009) and others (Hogervorst et al., 2008) havereported that athletes' cognitive performance – as assessed by theStroop test – is improved while exercising. However, the robustnessof this effect with aging and fitness status is unknown, nor are themediators of such effects. Performing aerobic exercise increases CBF(Hellstrom et al., 1996; Ogoh and Ainslie, 2009) as a consequence ofincreased brain neuronal activity and metabolism (Ide and Secher,
542 S.J.E. Lucas et al. / Experimental Gerontology 47 (2012) 541–551
2000; Secher et al., 2008). Such activation-flow coupling (termedneurovascular coupling) is a key mechanism for the maintenance ofadequate oxygen delivery to the brain (Girouard and Iadecola, 2006).Given the reported association between CBF and cognitive function atrest (Bertsch et al., 2009; Heo et al., 2010), one potential mechanismfor the improved cognitive performance while exercising is that it isrelated to changes in cerebral perfusion as a consequence of increasedneuronal activation.
The present study examined: i) whether increased cerebral perfu-sion while exercising is directly related to improved cognitive perfor-mance (executive function; as assessed by a Stroop task (Stroop,1935)); and (ii) how aging affects exercise-induced changes in cogni-tive performance. We hypothesized that cognitive performancewould be improved while exercising for both age groups, and theextent of improvement would be related to elevations in CBF. To testthis hypothesis, we combined transcranial Doppler to assess middlecerebral artery blood flow velocity (MCAv; index of CBF) and near infra-red spectroscopy (NIRS) during rest and exercisewith andwithout cog-nition testing. Transcranial Doppler provides a direct measure of bloodflow velocity, whereas NIRS provides activation-dependent informa-tion. The combination of NIRS and MCAv provides complementaryinformation for the evaluation of cerebral hemodynamic function andthis combination of methods has not previously been applied in studiesof healthy younger and older age groups during exercise and cognitiveperformance testing.
2. Methods
2.1. Participants
Thirteen young individuals (7 males, 6 females; aged 24±5 years;_VO2 max 32±6 mL·kg−1·min−1) and nine older individuals (4 males,5 females; 62±3 years; _VO2 max 24±5 mL·kg−1·min−1) volunteeredfor this study, which was approved by the University of Otago's HumanEthics Committee and conformed to the standards set by theDeclarationof Helsinki. Participants were informed of the experimental proceduresand possible risks involved in the study, and written informed consentwas obtained. All participants underwent full medical screening priorto inclusion in the study, including 12-lead ECG assessment. All partici-pants were inactive, participating in less than 30 min of exercise on lessthan three times per week for ≥2 years. Participants were not takingany medication, all were non-smokers, and none had any history of
Fig. 1. Schema representation of the experiment protocol. Thirteen younger and nine olderseated upright while resting, and the first simple and then most difficult blocks of the Stroopblood flow velocity (Doppler) and cerebral hemodynamics (near infrared spectroscopy) we
cardiovascular, cerebrovascular or respiratory disease. All older femaleswere post-menopausal and all young females were tested during theearly follicular phase of the menstrual cycle or during the sugar pillphase if they were taking the oral contraceptive pill. Participants werenot offered any financial reimbursement to be involved in the studyand recruitment was primarily through advertisement in the localmedia.
2.2. Experimental method and design
Prior to the experimental testing, all participantswere fully familiar-ized with all experimental procedures. Participants were asked toabstain from caffeine, alcohol and strenuous exercise in the 12 h priorto each testing session, and to avoid consuming a large meal in the 2 hprior to testing. Following successful screening into the study, partici-pantswere required to visit the laboratory on three occasions: 1) amax-imal aerobic power test ( _VO2max); 2) a full familiarization trial, and3) the data collection trial. As illustrated in Fig. 1, at the familiarizationand data collection trial sessions, following instrumentation, partici-pants sat upright for resting measures before completing two 8-minbouts of cycling exercise on a cycle ergometer (Excalibur Sport, LodeB.V., The Netherlands) at 30% followed by 70% heart rate range (HRR;established from the _VO2max test). Following baseline measures atrest and during exercise, participants' cognitive performance wasassessed using a computer-adapted Stroop test (described below).
2.3. Measurements
2.3.1. Cognitive performance: Adapted Stroop taskCognitive performance was assessed for both speed and accuracy
on a Stroop task (Stroop, 1935), which is a well-known paradigmfor investigating aspects of cognitive performance that depend on ex-ecutive functioning. Specifically, the task assesses cognitive processesassociated with selective attention to specific information and inhibi-tion of prepotent responses during decision-making tasks involvingstimuli and responses (see MacLeod, 1991 for detailed review). TheStroop task involves presenting color words (e.g., RED) displayed ineither congruent (red) or incongruent (e.g., green) colors, and controltrials in which only one dimension of the stimulus is present (i.e., pre-senting all words in black colors, or presenting color patches butwithout words; Stroop, 1935). The present study employed modifiedversions of the standard Stroop task to produce task blocks with
participants completed four blocks (ranging from simple to difficult) of the Stroop testtest while cycling at 30% and 70% of their heart rate range (HRR). Measures of cerebralre collected throughout the protocol (see text for full details).
543S.J.E. Lucas et al. / Experimental Gerontology 47 (2012) 541–551
increasing levels of difficulty so that performance measures on simpleand more cognitively-demanding tasks could be directly compared.The test involved either four (at rest) or two (during exercise) blocksof either 40 (rest) or 20 (exercise) trials in which the participantresponded by pressing the ‘z’ or ‘/’ key of a laptop computer, withthese keys corresponding to a response option at the bottom left (z)or right (/) of the screen (as shown in Fig. 2). The initial two blocks(panels b and d, Fig. 2) measured, respectively, response time to iden-tify the color name of a word written in a neutral color (i.e., wordreading), which was the simplest task condition, and the color of arectangle without any color-word information present (the next eas-iest condition). Blocks 3 and 4 were progressively more difficult, bothrequiring participants to respond on the basis of the color of the textbut while ignoring or inhibiting the prepotent response to the wordname. As shown in panel f, Fig. 2, the third block had response optionsprinted in a neutral color, whereas the fourth block (panel h) had theadditional source of difficulty in which the response choices furtherincreased the demands of the task, requiringparticipants again to inhibita prepotent word name. Thus, the blocks progressively increased indifficulty from1 to 4, by requiringparticipants to increase their demandsof selective attention (MacLeod, 1991). Participants were instructed to“attempt each trial as quickly and accurately as possible”. Each block
Fig. 2. An example of a trial in each of the four blocks in the computer-adapted Stroop tesexamples f) and h) represent more difficult choice response time tasks. The instructions fothe ‘z’ or ‘/’ key of a laptop computer, with these keys corresponding to a response option
began with a display of the instructions (see Fig. 2), which were alsogiven verbally to the participant. The exercise test consisted of blocks 1and 4 only (panels b and f, respectively, Fig. 2) to limit the time requiredto assess cognitive function at each intensity and thus ensure that allparticipants could complete both exercise intensities. Stimulus pre-sentation and response collection were performed using in-housealgorithms (MatLab v7), which included a randomization of trial pre-sentation order within each level of task difficulty. Participants werefully familiarized with the Stroop task, both at rest and while exercis-ing during the familiarization trial, and completed a shortened version(10 trials) of all blocks of the test before the resting condition of thedata collection trial.
2.3.2. Cerebral blood flow velocity and cortical hemodynamicsBlood flow velocity in the right middle cerebral artery (MCAv) was
measured using a 2-MHz pulsed Doppler ultrasound system (DWLDoppler, Compumedics Ltd, Germany) using search techniques de-scribed elsewhere (Aaslid et al., 1982; Willie et al., 2011). The Dopplerprobe was secured with a plastic headband device to maintain optimalinsonation position and angle throughout the protocol. Frontal corticalhemodynamics were monitored non-invasively on the right side ofthe forehead using NIRS [NIRO-200; Hamamatsu Photonics KK;
t (right side). Examples b) and d) represent simple choice response time tasks, whiler each level were displayed at the start of each block (left side). Participants presseddisplayed at the bottom left (z) or right (/) of the screen.
544 S.J.E. Lucas et al. / Experimental Gerontology 47 (2012) 541–551
Hamamatsu, Japan]. The methodology of this system has been de-scribed previously (Al-Rawi et al., 2001). In brief, the NIRS techniqueis based on the relative transparency of human tissue to near-infraredlight and on the oxygenation-dependent light absorption changescaused by the chromophores oxyhemoglobin (O2Hb) and deoxyhemo-globin (HHb). The NIRO-200 measures the absolute concentration(micromoles per liter) changes of O2Hb, HHb and total hemoglobin(tHb) based on an application of a modified Beer–Lambert law (Al-Rawi et al., 2001). Thus, although it is not possible to distinguishbetween them, this method measures the balance of oxygen supplyand oxygen demand directly at a capillary level in the cerebral corticaltissue (Mehagnoul-Schipper et al., 2000). All three measured parame-ters (O2Hb, HHb and tHb) are reported, since they contribute importantinformation both individually and collectively (Ekkekakis, 2009).Thetotal cortical oxygenation index (TOI%=O2Hb/tHb×100) is alsoreported, and was calculated by the NIRS system (Spatially ResolvedSpectroscopy method) from the light attenuation slope along the dis-tance from the emitting point as detected by two photodiodes in thedetection probe. The probes were housed in an optically dense plasticholder secured on the skin with tape to minimize extraneous light.The frontal cortical hemodynamic measures were obtained simulta-neously every 1 s throughout the experiment and are expressed as themagnitude of the change from the resting baseline value.
2.3.3. Respiratory gas exchangeParticipants breathed through a leak-free respiratory mask (Hans-
Rudolph 7900 series, Kansas City, MO, USA) attached to a Y-shapedtwo-way non-rebreathing valve (Hans-Rudolph 7900). The pressureof end-tidal CO2 (PETCO2) was sampled from the leak-free mask andmeasured using a fast responding gas analyzer (model CD-3A, AEITechnologies, Pittsburgh, PA, USA). The gas analyzer was calibratedusing a known concentration of CO2 prior to each testing session.
Cerebrovascular hemodynamics and respiratory variables were allacquired continuously at 200 Hz using an analog-to-digital converter(Powerlab/16SP ML795; ADInstruments) interfaced with a computer,and were subsequently analyzed using commercially available soft-ware (Chart v7.2, ADInstruments).
2.4. Data analysis
For each participant, mean response time and error rate were cal-culated for each block of trials. Means (± SD) for younger and oldergroups were then collated for each level of task difficulty. For restingmeasures, results for the simplest task (block 1) and most difficulttask (block 4) are reported because those best represented the ex-tremes of performance (simple versus most difficult) that we aimedto capture in this study. Moreover, only those two blocks were testedduring exercise; thus, only the simplest and most difficult levels ofthe task could be used to assess any exercise-related effect.
To allow for potential differences in the blood flow profile betweenrest and exercise, as well as between age groups, mean MCAv at restand during exercise was calculated using the integrated cardiac cyclicmean. Twominutes of baseline data were collected at rest. A 30-s aver-age was used for exercise baselines, after at least 90 s to allow steadystate to be obtained. Following the baseline period, cerebrovascularand respiratory data were averaged over the final 30 s (rest) or 15 s(exercise) of each level of the Stroop test.
2.5. Statistical analysis
A repeated-measures ANOVA (SPSS v17) was used to assess therelation between ‘task difficulty’ (3 levels; baseline, simple and diffi-cult levels of Stroop task) and concurrent exercise (3 levels; rest,30% and 70% HRR) for each dependent variable. Pairwise comparisons(Bonferroni adjusted) were applied to further evaluate specific influ-ences of significant main effects and interactions. The influence of age
was examined as a between-participants factor. Data are presented asmean±SD and statistical significance was accepted at an α-level of0.05. A linear regression was used to test the relation between MCAvand response times at rest and during exercise, as well as to test therelation between changes in MCAv and cortical hemodynamics (tHband TOI) with changes in response time during exercise.
3. Results
3.1. Rest
At rest, response times for the simple and difficult tasks were fas-ter for the younger group than the older group, by 151 ms (~20%) and446 ms (~40%), respectively (Pb0.01, Table 1). The MCAv was 32%higher in the younger compared to the older group at baseline (66vs. 50 cm·s−1, respectively; Pb0.01, Fig. 3A). The MCAv respondeddifferently to the Stroop task between age groups (P=0.02); it wasunchanged in the younger group (P=0.23) but decreased in theolder group during the difficult level of the cognitive task (−10% vs.simple; Pb0.01). This differential age effect was also observed in thePETCO2 responses (P=0.04), with no change across task difficultyin the younger group (P=0.24) and decreasing in the older groupduring the difficult level of the task (−3 mm Hg vs. baseline;P=0.05; Fig. 3B). Higher MCAv was related to faster response timesfor both simple (r=−0.69, R2=0.47) and difficult (r=−0.69,R2=0.47) levels of the Stroop task (Fig. 6A).
Cortical hemodynamic changes in [HHb] and [O2Hb] from baselineduring the Stroop task performance were similar between age groups([HHb], P=0.23; [O2Hb], P=0.70). For example, [HHb] decreasedfrom baseline values in both groups during the simple Stroop task(−0.5±0.7 μM, P=0.03) and decreased to a larger extent duringthe difficult task (−1.0±1.1 μM, Pb0.01). In contrast, [O2Hb] in-creased from baseline during the difficult task (+1.4±2.2 μM,Pb0.01; Fig. 4A and B). The counterbalanced nature of these changeswas such that tHb and total cortical oxygenation (as indexed by TOI)were not altered from baseline measures (task difficulty main effect:P=0.13 and P=0.67, tHb and TOI respectively) for either age group(task difficulty×age effect: P=0.55 and P=0.25, respectively;Figs. 4C and 5).
3.2. Exercise
The difference in response times between the groups was notaltered with exercise (P=0.76); response times for the simple Strooptask were faster in the younger group compared to the older group by198 ms (31%) at the 30% HRR exercise intensity and by 152 ms (23%)at the 70% HRR exercise intensity. For the difficult Stroop task, re-sponse times were faster in the younger group by 374 ms (35%) at30% HRR exercise intensity and by 393 ms (40%) at the 70% HRR exer-cise intensity (Table 1). Response times on the simple task werenot reliably different during exercise compared to rest (P=0.12;Table 1), and this was common to both age groups (P=0.40). In con-trast, response times on the difficult task improved during exercise(Pb0.01), for both younger and older groups alike (group effect:P=0.42), and a differential effect of exercise intensity was also ap-parent (P=0.04, Table 1). Specifically, response times were ~70 msfaster when exercising at 70% compared to 30% HRR.
The MCAv remained ~32% higher in the younger compared to theolder group during exercise (76 vs. 58 cm·s−1 and 73 vs. 54 cm·s−1,at 30 and 70% HRR exercise, respectively; Pb0.01, Fig. 3A). Thus,exercise-related elevations in MCAv were not different betweengroups (P=0.49) or between the two intensities (P=0.22, Fig. 3A).The MCAv during cognitive testing was not altered compared to base-line measures (P=0.56), nor was it differentially affected by age(group effect: P=0.53; Fig. 3A). End-tidal PCO2 was progressivelyelevated during exercise compared to rest (Pb0.01, Fig. 3B) in both
Table 1Mean±SD response times (ms) and error rate (per 40 or 20 trials, rest and exercise respectively) for simple and difficult Stroop task performance in younger (n=13) and older(n=9) participants at rest and during 30% and 70% heart rate range (HRR) cycling exercise.
⁎ Different compared to younger group (Pb0.05).† Different compared to rest (Pb0.05).β Different compared to 30% HRR (Pb0.05).
545S.J.E. Lucas et al. / Experimental Gerontology 47 (2012) 541–551
groups (P=0.49), and was unchanged from baseline during cognitivetesting at each intensity (P>0.05; Fig. 3B).
The baseline [HHb] increased during exercise (Pb0.01 vs. rest), andby a similar extent for the two age groups (exercise×age: P=0.31).On average, [HHb] at 30% HRR increased from rest by 2.2±1.7 μM(Pb0.01). It was reduced at 70% HRR compared to 30% HRR exercise(P=0.03), but was still higher than at rest (change from rest: +1.2±1.2 μM; Pb0.01; Fig. 4A). The pattern of response during Stroop taskperformance was similar in both groups at both exercise intensities(P=0.37), but was differentially affected by exercise intensity(Pb0.01). Specifically, whereas [HHb] during simple task performancewas unchanged compared to baseline at both exercise intensities,[HHb] decreased while completing the difficult task during 30% HRR
Fig. 3. Cerebral blood flow (indexed by MCAv, A) and partial pressure of end tidal CO2 (PETCStroop task performance in younger (left, n=13) and older (right, n=9) participants whil(Pb0.05) compared to younger group; † different compared to rest; ‡ different compared toto 30% HRR exercise. Data are mean±SD.
exercise (by ~0.4 μM, P≤0.01), but was unchanged from baseline andsimple task responses during 70% HRR exercise (P>0.15; Fig. 4A).
The baseline [O2Hb] response to exercise intensity was differentbetween age groups (P=0.02; Fig. 4B). Baseline [O2Hb] at 30% HRRexercise tended to decrease in the younger group (change fromrest: −2.5±3.6 μM; P=0.08), while [O2Hb] was unchanged fromrest in the older group (P=1.00). In contrast, [O2Hb] at 70% HRRincreased relative to both resting and 30% HRR exercise baselines(Pb0.05) for both groups [change from rest: +5.3±3.9 μM (younger)and +3.7±2.8 μM (older)]; moreover, there was a greater increase in[O2Hb] for the younger group between 30 and 70% HRR exercise(P=0.02; Fig. 4B). The pattern of response during Stroop task perfor-mance was similar for both groups and exercise intensities (P=0.34).
O2, B) at baseline (white bars) and during simple (gray bars) and difficult (black bars)e at rest and during cycling exercise at 30% and 70% heart rate range (HRR). * differentsimple level of the Stroop task; α different compared to baseline; β different compared
Fig. 4. The absolute change from baseline at rest in deoxygenated hemoglobin (HHb, A), oxygenated hemoglobin (O2Hb, B) and total hemoglobin (tHb, C) during simple (gray bars)and difficult (black bars) Stroop task performance in younger (left, n=13) and older (right, n=9) participants while at rest and during cycling exercise at 30% and 70% heart raterange (HRR). White bars show change from rest for baseline measures at 30 and 70% HRR; * different (Pb0.05) compared to younger group; † different compared to rest; ‡ differentcompared to simple level of the Stroop task; α different compared to baseline; β different compared to 30% HRR exercise. Data are mean±SD.
546 S.J.E. Lucas et al. / Experimental Gerontology 47 (2012) 541–551
For example, [O2Hb] while exercising increased by 1.7±2.3 μM andby 1.1±1.4 μM from baseline and simple task responses, respectively,during performance of the difficult task (Pb0.01; Fig. 4B).
The baseline [tHb] response to exercise intensity was differentbetween age groups (P=0.01; Fig. 4C). Baseline [tHb] at 30% HRRexercise was unchanged compared to rest for both groups (P>0.12).In contrast, [tHb] increased more between 30 and 70% HRR for theyounger group than for the older group (change from rest: +6.3±3.9 μM for the younger group, and +5.1±2.6 μM for the older group,producing a significant age×exercise effect: P=0.01; Fig. 4C). The pat-tern of response during Stroop task performance was similar for bothgroups and exercise intensities (P=0.19). Specifically, [tHb] whileexercising increased by 1.6±2.1 μM and by 1.0±1.4 μM from baselineand simple task responses, respectively, during performance of thedifficult task (Pb0.01; Fig. 4C).
Calculated total cortical oxygenation (i.e., TOI) decreased similarlyfor both groups (P=0.25) at 30% HRR exercise baseline compared to
resting and 70% HRR baseline (P≤0.01; Fig. 5). The pattern of thisresponse during Stroop task performance was similar in both groupsat both exercise intensities (P=0.65), but was differentially affectedby exercise intensity (Pb0.01). Stroop task performance did not reli-ably affect TOI at 30% HRR exercise (P=0.09), whereas TOI decreasedduring the simple and difficult Stroop task performance at 70% HRR(both Pb0.05 vs. baseline, Fig. 5).
3.3. Correlations between cerebral perfusion and response times duringexercise
The evident relation between absolute MCAv and simple and diffi-cult response times at rest (R2=0.47 and 0.47, respectively) was re-duced at 30% HRR exercise (R2=0.26 and 0.29, respectively) and wasfurther attenuated at 70% HRR exercise (R2=0.13 and 0.17, respec-tively; Fig. 6B and C). The correlations between exercise-inducedchanges in response time on the difficult task and changes in MCAv
Fig. 5. The calculated total cortical oxygenation (A), and change from baseline (B), during simple (gray bars) and difficult (black bars) Stroop task performance in younger (left,n=13) and older (right, n=9) participants while at rest and during cycling exercise at 30% and 70% heart rate range (HRR). White bars show absolute values for baseline measures(A) and change from rest for baseline measures at 30 and 70% HRR (B); * different (Pb0.05) compared to younger group; † different compared to rest; ‡ different compared tosimple level of the Stroop task; α different compared to baseline; β different compared to 30% HRR exercise. Data are mean±SD.
547S.J.E. Lucas et al. / Experimental Gerontology 47 (2012) 541–551
were negligible (absolute, R2=0.06; relative, R2b0.01) as were cor-relations between changes in response time to the difficult task andchanges in [tHb] (R2=0.07) or the calculated total cortical oxygena-tion index (R2=0.04).
4. Discussion
The main findings of this study were that: 1) despite persistentage-related differences in cognitive performance (executive function-ing; as assessed by the Stroop task) at rest and during exercise, therewas a progressive improvement during exercise in both younger andolder participants; 2) a strong relation between absolute MCAv andresponse times on the Stroop tasks was evident at rest, which wasattenuated during each progressive elevation of exercise intensity,and 3) global CBF was dissociated from regional cortical oxygenationand blood volume during exercise and Stroop task performance. Col-lectively, these data indicate that regardless of age, executive functioncognition while exercising is improved compared to the resting state.Furthermore, while global CBF is strongly related to this type of cog-nitive processing at rest, this relation becomes uncoupled duringexercise. Finally, regulatory mechanisms appear to alter regional cor-tical oxygenation and blood volume, at least to the prefrontal cortex,independently of global CBF regulation during both exercise and per-formance of the Stroop task.
4.1. Methodological considerations
4.1.1. Assessment of cerebral perfusionWhile caution should to be taken when interpreting absolute CBF
values from transcranial Doppler technology, previous reports indi-cate that MCAv is a reliable index of global cerebral blood flow bothat rest and during exercise (reviewed in: Ogoh and Ainslie, 2009). Im-portantly, the current data are consistent with others that havereported age-related differences in MCAv (Ainslie et al., 2008; Buijs
et al., 1998), and the progressive decline in CBF from the age of 30to 70 years has been confirmed repeatedly using a variety of imagingtechniques (Beason-Held et al., 2007; Buijs et al., 1998; Kashimadaet al., 1994; Scheel et al., 2000; Stoquart-ElSankari et al., 2007).
Although NIRS assessment of cerebrovascular function is limitedto relatively superficial layers of tissue and has poorer spatial resolu-tion compared to other neuroimaging techniques (e.g., PET scanning,fMRI), the high temporal resolution, non-invasive measurement pro-cedure, relatively low cost, lack of interference from electromagneticradiation and an acceptable signal-to-noise ratio during exercise (inaddition to allowance for movement that is not possible in thefMRI/PET scanner due to problems in movement artifacts) makesNIRS the only viable technique to measure cortical activity during ex-ercise (reviewed in: Ekkekakis, 2009). Moreover, a recent study hasdemonstrated thatNIRS provides ametric of cognitive activation similarto fMRI during cognitive task performance (Cui et al., 2011). Anotherkey advantage of NIRS is that observed changes can be attributed to un-derlying physiological causes with a high degree of specificity (Obrigand Villringer, 2003). To achieve the greatest level of spatial resolutionpossible with this technique the detector probes were separated by5 cm, allowing the light penetration depth to approximately 2.5 cm(Ferrari et al., 2004). Furthermore, it has also been reported thatextra-cranial contribution to the NIRS signal is negligible when theinteroptode distance is >4.5 cm (Smielewski et al., 1995). Finally, toaddress the limitations of the NIRS technology, we have restricted ourinterpretations to changes from baseline resting measures, as recom-mended by others (Subudhi et al., 2009).
4.1.2. Assessment of cognition with the Stroop taskThe Stroop task is a well-used paradigm for investigating aspects
of cognitive performance that depend on executive functioning. Theexecutive control processes, and the brain regions that supportthem (chiefly the prefrontal cortex), show substantial age-related de-terioration (Hillman et al., 2008). Moreover, exercise training has
Fig. 6. Relation between absolute cerebral blood flow (MCAv, cm·s−1) and response time for simple (left column) and difficult (right column) tasks as assessed by a computer-adapted Stroop test in younger (closed circle; n=13; 24±5 y) and older (open circle; n=9; 62±3 y) participants at (A) rest during (B) 30% and (C) 70% heart rate range(HRR) cycling exercise. Each point represents an individual value. These data illustrate that the strong relation between absolute MCAv and response time at rest was attenuatedduring each progressive elevation of exercise intensity.
548 S.J.E. Lucas et al. / Experimental Gerontology 47 (2012) 541–551
been shown to affect executive control processes (Colcombe andKramer, 2003). For these reasons we view the Stroop task as a mean-ingful and practical measure to assess age-related and exercise-induced differences in cognition.
4.1.3. Cross-sectional study designOur study design was cross-sectional in nature. Thus, we cannot
exclude the influence of selective mortality or morbidity and cohortdifferences accounting for some of the age effects we have observedhere. However, given that improved cognitive performance whileexercising did not differ between the two age groups, this concern
regarding cross-sectional study design would appear less relevant.Finally, we intentionally selected healthy older (and younger) partic-ipants for this study; therefore the findings of the present study canonly be inferred to healthy older adults. Further study is needed to es-tablish whether disease populations will also benefit from exercisewhile performing cognitive tasks.
Finally, we acknowledge that 90 s is a relatively short period toallow for an exercising steady state to be achieved. Nevertheless, fol-lowing the cognitive testing we measured a second baseline period[used for cerebrovascular responsiveness measures (reported else-where, Murrell et al. in press)] and subsequent analysis showed that
549S.J.E. Lucas et al. / Experimental Gerontology 47 (2012) 541–551
there were no differences between the two exercise baselines (30%HRR, P=0.79; 70% HRR, P=0.55). Thus, we are confident that the ex-ercise baselines reported here represent a steady state.
4.2. Cognitive function and cerebral perfusion at rest
The observed difference between younger and older participantsin performance on both simple and difficult cognitive tasks in the pre-sent study is consistent with literature on age-related cognitive de-cline (Craik and Salthouse, 2000; Park and Schwarz, 2000). Indeed,a consistent decline in cognitive performance for a wide variety oftasks across the lifespan (3rd to 9th decade) has been elegantly sum-marized (Park et al., 2001). Our data indicate that approximately halfof the increase in response times for both simple and difficult taskswas accounted for by reductions in absolute MCAv (as an index ofglobal CBF (Secher et al., 2008); Fig. 6A). Age-related reductions inCBF most likely reflect a reduction in brain volume (i.e., global atrophy)because less neuronal demand requires less CBF. However, it remainsunknown whether chronically-reduced flow drives atrophy and thusreduced demand (de la Torre, 2010), or vice versa, or both. Global atro-phy, especially in the frontal lobe and hippocampus, has been suggestedto be the cause of age-related cognitive decline (Kramer et al., 2007),which has also been shown to be related to reduced CBF (Bertschet al., 2009; Heo et al., 2010). Indeed, lower CBF in clinical populationswith known impaired cognitive function (e.g., dementia (Lange-Asschenfeldt and Kojda, 2008)) further supports the influence of CBFon cognitive function. Furthermore, Marshall et al. (2001) observedimpaired cognition (attention) after lowering CBF (by ~40%; similar tothe cross sectional observations in the present study) using internalcarotid artery balloon occlusion, an effect which was reversed whenthe occlusion was removed and CBF returned to normal. Thus, CBF isclearly important for optimal cognitive function, both indirectly (as aconsequence and/or cause of larger brain volume) and directly (viarestricted neuronal blood flow). Our data obtained at rest support thenotion that higher blood flow, as indexed by flow velocity, is associatedwith better cognitive performance in humans. Therefore, any interven-tions (e.g., regular aerobic exercise training) that improve resting CBF(Ainslie et al., 2008; Rhyu et al., 2010) may also offset the age-relateddecline in cognitive function.
In addition, we observed a further widening of the difference inresting MCAv when the older participants were completing theStroop task (Fig. 3A). The corresponding drop in PETCO2 (Fig. 3B)would indicate mild hyperventilation in these older participants,which would reduce arterial PCO2 and induce a 2–3% decline inMCAv per millimeter of mercury (Peebles et al., 2007). Thus, theolder participants' age-related difference in MCAv, as observed inbaseline measures, was made worse by their ‘ventilation strategy’while performing our difficult cognitive task. Such hyperventilationmay influence cognitive performance indirectly via its direct effect onCBF. The reason for this ventilation strategy is not clear, but appearedto be consistent within the older group (i.e., no outlier effect). Notably,most of the older participants in the present study utilize computersregularly and all participants were fully familiarized with the task,so it is unlikely that this is related to less familiarity with computers.
4.3. Cognitive function and cerebral perfusion during exercise
The positive chronic effects of exercise on brain function are be-coming widely acknowledged (e.g., Voss et al., 2011), however littlehas been reported on exercise-induced changes in cerebral perfusionand any direct effect on cognitive performance during an individualbout of exercise. Furthermore, no studies prior to the present onehave examined the age effect. Thus, the observed effects of age oncognitive performance during exercise in the present study arenovel. Two key findings of this study were that MCAv in both youngerand older participants increased similarly during exercise; and that
there was a reduction in the neurovascular coupling regulation ofCBF with progressive exercise intensity (Fig. 3). It seems unlikelythat exercising hyperventilation-induced hypocapnia explains thisuncoupling, as has been shown during exhaustive exercise (Ogoh etal., 2005), given that PETCO2 was unchanged between our two sub-maximal exercise intensities (Fig. 3B). Whatever the cause, it wouldalso appear not to be affected by age, given that the pattern ofexercise-induced elevations in MCAv was similar for both youngerand older participants. In addition, our findings show that thisexercise-induced improvement in Stroop task performance alsooccurs in untrained individuals with low fitness status ( _VO2max of25 and 32 mL·kg−1·min−1 in our older and younger participants, re-spectively) as well as in athletes that we (54 mL·kg−1·min−1, Lucaset al., 2009) and others (57 mL·kg−1·min−1, Hogervorst et al., 2008)have previously reported on. We hypothesized that the exercise-induced increase in CBF would be the mechanism for the exerciseeffect on Stroop task performance. However, while we did observean improvement in response times during exercise, this improvementwas apparently not related to changes in MCAv (or total hemoglobin(tHb) or the calculated cortical oxygenation (i.e., TOI) sampled at theprefrontal cortex). This finding is further illustrated in progressivereductions in the relation between absolute MCAv and responsetime on the Stroop task during the two exercise intensities comparedto rest (Fig. 6). Thus, it would seem that other factors that are lessaffected by age might explain exercise-related improvements in cog-nition. Given the differential effect on response time with exerciseintensity, neural or humoral (e.g., SNS and/or noradrenaline) or met-abolic (e.g., greater oxygen extraction, or Q10 effect of temperature)factors would seem the most likely candidates. Further studies exam-ining the causation of concurrent (as well as chronic) exercise-relatedimprovements in cognition seem important.
The activation-dependent information obtained from the corticalhemodynamic responses provides some further intriguing observa-tions. Consistent with the observations of others (Ide et al., 1999;Subudhi et al., 2007; Subudhi et al., 2009), [HHb], [O2Hb] and [tHb]increased during exercise as a consequence of the elevated CBF asso-ciated with increased neuronal activity and metabolism with exercise(Ide and Secher, 2000; Secher et al., 2008); although this was not uni-versal (e.g., younger [O2Hb], Fig. 4B). Moreover, we observed a differ-ential effect of intensity for [O2Hb] and [tHb] responses and an age-related difference in these increases, neither of which were evidentin the MCAv responses. Thus, changes in MCAv and tHb were not astightly coupled as has been reported previously at similar exerciseintensities (Ide et al., 1999; Subudhi et al., 2007); although in a laterstudy by Subudhi et al. (2008), an uncoupling of the MCAv and tHbrelation was reported, albeit only at higher exercise intensities. In-deed, given that both [O2Hb] and [tHb] are thought to reflect changesin regional cerebral blood volume (Obrig and Villringer, 2003), ourdata are consistent with animal-based studies that report regionalCBF differences during exercise (Delp et al., 2001; Gross et al.,1980), as well as indicate that regional CBF to the prefrontal cortexin humans is reduced during exercise with aging.
Another novel observation was the regional CBF changes to theprefrontal cortex during Stroop task performance (Fig. 4). Our dataindicate that regional cerebral blood volume (as indexed by [O2Hb]and [tHb]) was increased during a cognitive task that required exec-utive functioning. Our findings are consistent with previous findingsin humans at rest, reporting increased regional CBF to areas associatedwith elevated neuronal activity during cognition (Risberg and Ingvar,1973). The present study is the first to show that this response is pre-served during exercise, given that the pattern of response for [HHb],[O2Hb] and [tHb]was similar at rest and during both exercise intensitiesregardless of the resting and exercise baseline values. Furthermore,these regional CBF changes to the prefrontal cortex occur indepen-dently of blood flow in the MCA (Fig. 3A). Finally, the age-related dif-ferential increases in regional cerebral blood volume increases
550 S.J.E. Lucas et al. / Experimental Gerontology 47 (2012) 541–551
between 30% and 70% HRR exercise intensities may explain the greater,on average, oxygen extraction (as reflected by the change in the totalcortical oxygenation index) in the older group during Stroop task per-formance at 70% HRR (Fig. 5). Overall, these NIRS data indicate that re-gional shifts in cerebral blood volume, independent ofMCAflow, supplyprefrontal regions of the brain during exercise and during cognitive per-formance. Nevertheless, even though these cortical hemodynamicresponses are different with aging, the overall age-effect on cognitiondid not change, nor did the cortical hemodynamic responses explainthe exercise-induced improvements in cognitive performance for bothgroups. It is important to note thatwe onlymeasured the right prefron-tal cortex, sincewewanted to contrast this region directlywith the rightMCAv measures. It has been reported that due to the verbal nature ofthe Stroop task it is left hemisphere lateralized (Nee et al., 2007), andconsistently stronger effects have been reported on the left comparedto the right prefrontal cortex (e.g., Yanagisawa et al., 2010). Our findingsare therefore conservative and indicate a generality of findings even tothe less dominant hemisphere involvement of prefrontal activity. Asmentioned above, a central question for the current study was to com-pare global and regional (prefrontal) blood flow during exercise, whichis clearly captured by our right hemisphere measures.
4.4. Implications
A direct and novel implication of the present findings is that bothyounger and older participants can expect improvements in cognitionwhile performing exercise. While we cannot explain the mechanismsfor the improvement, the current study illustrates that not only willolder people benefit with habitual exercise programs which havebeen shown to elevate CBF, but they might expect the same enhancedcognitive (executive) function as their younger counterparts whileperforming exercise. The extent to which these changes are main-tained following exercise warrants further study.
An indirect implication is that exercise, when undertaken regularly,stands to have much more beneficial impact on cognition (and overallhealth) in older adults, when resting (Cotman and Berchtold, 2002;Hillman et al., 2008; Voss et al., 2011). Exercise promotes extensive vas-cular changes and adaptive mechanisms in the central nervous system,such as the induction of vascular antioxidant pathways, capillarygrowth, neurogenesis and enhanced synaptic plasticity (Churchill etal., 2002; Vaynman and Gomez-Pinilla, 2006), consequently improvingcerebral perfusion via increased brain volume and metabolism(Colcombe et al., 2006; Pantano et al., 1984). Such vascular and neuraladaptations have been suggested to improve and maintain cognitivefunction during aging (Lange-Asschenfeldt and Kojda, 2008). Further-more, physical activity has been described as having a protective effectfor cognitive decline or dementia (Deary et al., 2006; Larson et al., 2006;Tabbarah et al., 2002). Exercise may increase brain reserve by promot-ing healthy cardiovascular function and diminishing cerebrovasculardisease burden (Shinton and Sagar, 1993). Studies in animals haveshown that exercise slows the expression of Alzheimer disease-likepathology (Adlard et al., 2005) and enhances brain health and functionby priming molecular memory for the plasticity molecular brain-derived neurotrophic factor (Berchtold et al., 2005; Hillman et al.,2008). Exercise may thus become a powerful tool in the prevention ofnot only vascular cognitive impairment but also of neurodegenerativedisorders such as Alzheimer's disease, the most common form ofdementia (Lange-Asschenfeldt and Kojda, 2008). Further research isurgently needed in humans to better define the use of exercise as a prin-cipal means by which to improve CBF and thus offset age-relateddeclines in cognitive function (when exercising or especially resting).
Author contributions
All the listed authors were involved in: 1) the conception anddesign, or analysis and interpretation of data; 2) drafting the article
or revising it critically for important intellectual content, and 3) finalapproval of the version to be published.
Grants
This study was supported by the Health Research Council (GrantNo: 06/230) and National Heart Foundation of New Zealand.
Conflicts of interest
None.
References
Aaslid, R., Markwalder, T.M., Nornes, H., 1982. Noninvasive transcranial Doppler ultra-sound recording of flow velocity in basal cerebral arteries. J. Neurosurg. 57, 69–74.
Adlard, P.A., Perreau, V.M., Pop, V., Cotman, C.W., 2005. Voluntary exercise decreasesamyloid load in a transgenic model of Alzheimer's Disease. J. Neurosci. 25,4217–4221.
Ainslie, P.N., Cotter, J.D., George, K.P., Lucas, S.J.E., Murrell, C., Shave, R., Thomas, K.,Williams,M.J.A., Atkinson, G., 2008. Elevation in cerebral blood flow velocity with aerobic fitnessthroughout healthy human ageing. J. Physiol. 586, 4005–4010.
Al-Rawi, P.G., Smielewski, P., Kirkpatrick, P.J., 2001. Evaluation of a near-infrared spec-trometer (NIRO 300) for the detection of intracranial oxygenation changes in theadult head. Stroke 32, 2492–2500.
Beason-Held, L.L., Moghekar, A., Zonderman, A.B., Kraut, M.A., Resnick, S.M., 2007. Lon-gitudinal changes in cerebral blood flow in the older hypertensive brain. Stroke 38,1766–1773.
Berchtold, N.C., Chinn, G., Chou, M., Kesslak, J.P., Cotman, C.W., 2005. Exercise primes amolecular memory for brain-derived neurotrophic factor protein induction in therat hippocampus. Neuroscience 133, 853–861.
Bertsch, K., Hagemann, D., Hermes, M., Walter, C., Khan, R., Naumann, E., 2009. Restingcerebral blood flow, attention, and aging. Brain Res. 1267, 77–88.
Buijs, P.C., Krabbe-Hartkamp, M.J., Bakker, C.J., de Lange, E.E., Ramos, L.M., Breteler,M.M., Mali, W.P., 1998. Effect of age on cerebral blood flow: measurement withungated two-dimensional phase-contrast MR angiography in 250 adults. Radiology209, 667–674.
Churchill, J.D., Galvez, R., Colcombe, S., Swain, R.A., Kramer, A.F., Greenough, W.T., 2002.Exercise, experience and the aging brain. Neurobiol. Aging 23, 941–955.
Colcombe, S., Kramer, A.F., 2003. Fitness effects on the cognitive function of olderadults: a meta-analytic study. Psychol. Sci. 14, 125–130.
Colcombe, S.J., Erickson, K.I., Scalf, P.E., Kim, J.S., Prakash, R., McAuley, E., Elavsky, S.,Marquez, D.X., Hu, L., Kramer, A.F., 2006. Aerobic exercise training increasesbrain volume in aging humans. J. Gerontol. A Biol. Sci. Med. Sci. 61, 1166–1170.
Cotman, C.W., Berchtold, N.C., 2002. Exercise: a behavioral intervention to enhancebrain health and plasticity. Trends Neurosci. 25, 295–301.
Craik, F.I.M., Salthouse, T.A. (Eds.), 2000. The Handbook of Aging and Cognition, 2nd ed.Lawrence Erlbaum Associates, Inc., Mahwah, NJ.
Cui, X., Bray, S., Bryant, D.M., Glover, G.H., Reiss, A.L., 2011. A quantitative comparisonof NIRS and fMRI across multiple cognitive tasks. NeuroImage 54, 2808–2821.
de la Torre, J.C., 2010. Vascular risk factor detection and control may prevent Alzhei-mer's disease. Ageing Res. Rev. 9, 218–225.
Delp, M.D., Armstrong, R.B., Godfrey, D.A., Laughlin, M.H., Ross, C.D., Wilkerson, M.K.,2001. Exercise increases blood flow to locomotor, vestibular, cardiorespiratoryand visual regions of the brain in miniature swine. J. Physiol. 533, 849–859.
Ekkekakis, P., 2009. Illuminating the black box: Investigating prefrontal cortical hemo-dynamics during exercise with near-infrared spectroscopy. J. Sport Exerc. Psychol.31, 505–553.
Ferrari, M., Mottola, L., Quaresima, V., 2004. Principles, techniques, and limitations ofnear infrared spectroscopy. Can. J. Appl. Physiol. 29, 463–487.
Girouard, H., Iadecola, C., 2006. Neurovascular coupling in the normal brain and in hyper-tension, stroke, and Alzheimer disease. J. Appl. Physiol. 100, 328–335.
Gross, P.M., Marcus, M.L., Heistad, D.D., 1980. Regional distribution of cerebral bloodflow during exercise in dogs. J. Appl. Physiol. 48, 213–217.
Hellstrom, G., Fischer-Colbrie, W., Wahlgren, N.G., Jogestrand, T., 1996. Carotid arteryblood flow and middle cerebral artery blood flow velocity during physical exercise.J. Appl. Physiol. 81, 413–418.
Hill, R.D., Storandt, M., Malley, M., 1993. The impact of long-term exercise training onpsychological function in older adults. J. Gerontol. 48, P12–P17.
Hillman, C.H., Erickson, K.I., Kramer, A.F., 2008. Be smart, exercise your heart: exerciseeffects on brain and cognition. Nat. Rev. Neurosci. 9, 58–65.
Hogervorst, E.E.F., Bandelow, S., Schmitt, J., Jentjens, R.O.Y., Oliveira, M., Allgrove, J.,Carter, T.O.M., Gleeson, M., 2008. Caffeine improves physical and cognitive perfor-mance during exhaustive exercise. Med. Sci. Sports Exerc. 40, 1841–1851.
551S.J.E. Lucas et al. / Experimental Gerontology 47 (2012) 541–551
Ide, K., Secher, N.H., 2000. Cerebral blood flow and metabolism during exercise. Prog.Neurobiol. 61, 397–414.
Ide, K., Horn, A., Secher, N.H., 1999. Cerebral metabolic response to submaximal exer-cise. J. Appl. Physiol. 87, 1604–1608.
Kashimada, A., Machida, K., Honda, N., Mamiya, T., Takahashi, T., Kamano, T., Inoue, Y.,Osada, H., 1994. Measurement of cerebral blood flow in normal subjects by phasecontrast MR imaging. Nippon Igaku Hoshasen Gakkai Zasshi 54, 1116–1125.
Kramer, A.F., Hahn, S., Cohen, N.J., Banich, M.T., McAuley, E., Harrison, C.R., Chason, J.,Vakil, E., Bardell, L., Boileau, R.A., Colcombe, A., 1999. Ageing, fitness and neurocog-nitive function. Nature 400, 418–419.
Kramer, J.H., Mungas, D., Reed, B.R., Wetzel, M.E., Burnett, M.M., Miller, B.L., Weiner,M.W., Chui, H.C., 2007. Longitudinal MRI and cognitive change in healthy elderly.Neuropsychology 21, 412–418.
Lange-Asschenfeldt, C., Kojda, G., 2008. Alzheimer's disease, cerebrovascular dysfunc-tion and the benefits of exercise: from vessels to neurons. Exp. Gerontol. 43,499–504.
Larson, E.B., Wang, L., Bowen, J.D., McCormick, W.C., Teri, L., Crane, P., Kukull, W., 2006.Exercise is associated with reduced risk for incident dementia among persons 65years of age and older. Ann. Intern. Med. 144, 73–81.
Lucas, S.J.E., Anson, J.G., Palmer, C.D., Hellemans, I.J., Cotter, J.D., 2009. The impact of100 hours of exercise and sleep deprivation on cognitive function and physicalcapacities. J. Sports Sci. 27, 719–728.
MacLeod, C.M., 1991. Half a century of research on the Stroop effect: an integrativereview. Psychol. Bull. 109, 163–203.
Marshall, R.S., Lazar, R.M., Pile-Spellman, J., Young,W.L., Duong, D.H., Joshi, S., Ostapkovich,N., 2001. Recovery of brain function during induced cerebral hypoperfusion. Brain124, 1208–1217.
Mehagnoul-Schipper, D.J., Vloet, L.C.M., Colier, W.N.J.M., Hoefnagels, W.H.L., Jansen,R.W.M.M., 2000. Cerebral oxygenation declines in healthy elderly subjects in re-sponse to assuming the upright position. Stroke 31, 1615–1620.
Murrell, C.J., Cotter, J.D., Thomas, K.N., Lucas, S.J., Williams, M.J., Ainslie, P.N., in press.Cerebral blood flow and cerebrovascular reactivity at rest and during sub-maximalexercise: effect of age and 12-week exercise training. Age. http://dx.doi.org/10.1007/s11357–012–9414-x.
Nee, D., Wager, T., Jonides, J., 2007. Interference resolution: insights from a meta-analysis of neuroimaging tasks. Cogn. Affect. Behav. Neurosci. 7, 1–17.
Obrig, H., Villringer, A., 2003. Beyond the visible— imaging the human brain with light.J. Cereb. Blood Flow Metab. 23, 1–18.
Ogoh, S., Ainslie, P.N., 2009. Cerebral blood flow during exercise: mechanisms of regu-lation. J. Appl. Physiol. 107, 1370–1380.
Pantano, P., Baron, J., Lebrun-Grandie, P., Duquesnoy, N., Bousser, M., Comar, D., 1984.Regional cerebral blood flow and oxygen consumption in human aging. Stroke15, 635–641.
Park, D.C., Polk, T.A., Mikels, J.A., Taylor, S.F., Marshuetz, C., 2001. Cerebral aging: inte-gration of brain and behavioral models of cognitive function. Dialogues Clin. Neu-rosci. 3, 151–165.
Peebles, K.C., Celi, L., McGrattan, K., Murrell, C.J., Thomas, K.N., Ainslie, P.N., 2007.Human cerebrovascular and ventilatory CO2 reactivity to end-tidal, arterial andinternal jugular vein PCO2. J. Physiol. 584, 347–357.
Rhyu, I.J., Bytheway, J.A., Kohler, S.J., Lange, H., Lee, K.J., Boklewski, J., McCormick, K.,Williams, N.I., Stanton, G.B., Greenough, W.T., Cameron, J.L., 2010. Effects of aerobicexercise training on cognitive function and cortical vascularity in monkeys. Neuro-science 167, 1239–1248.
Risberg, J., Ingvar, D.H., 1973. Patterns of activation in the grey matter of the dominanthemisphere during memorizing and reasoning. Brain 96, 737–756.
Scheel, P., Ruge, C., Petruch, U.R., Schoning, M., 2000. Color duplex measurement ofcerebral blood flow volume in healthy adults. Stroke 31, 147–150.
Secher, N.H., Seifert, T., Van Lieshout, J.J., 2008. Cerebral blood flow and metabolismduring exercise: implications for fatigue. J. Appl. Physiol. 104, 306–314.
Shinton, R., Sagar, G., 1993. Lifelong exercise and stroke. BMJ 307, 231–234.Smielewski, P., Kirkpatrick, P., Minhas, P., Pickard, J.D., Czosnyka, M., 1995. Can cere-
brovascular reactivity be measured with near-infrared spectroscopy? Stroke 26,2285–2292.
Stoquart-ElSankari, S., Baledent, O., Gondry-Jouet, C., Makki, M., Godefroy, O., Meyer,M.E., 2007. Aging effects on cerebral blood and cerebrospinal fluid flows. J. Cereb.Blood Flow Metab. 27, 1563–1572.
Stroop, J.R., 1935. Studies of interference in serial verbal reactions. J. Exp. Psychol. 18,643–662.
Subudhi, A.W., Dimmen, A.C., Roach, R.C., 2007. Effects of acute hypoxia on cerebral andmuscle oxygenation during incremental exercise. J. Appl. Physiol. 103, 177–183.
Subudhi, A.W., Lorenz, M.C., Fulco, C.S., Roach, R.C., 2008. Cerebrovascular responses toincremental exercise during hypobaric hypoxia: effect of oxygenation on maximalperformance. Am. J. Physiol. Heart Circ. Physiol. 294, H164–H171.
Subudhi, A.W., Miramon, B.R., Granger, M.E., Roach, R.C., 2009. Frontal and motor cor-tex oxygenation during maximal exercise in normoxia and hypoxia. J. Appl. Phy-siol. 106, 1153–1158.
Tabbarah, M., Crimmins, E.M., Seeman, T.E., 2002. The relationship between cognitiveand physical performance. J. Gerontol. A Biol. Sci. Med. Sci. 57, M228–M235.
Vaynman, S., Gomez-Pinilla, F., 2006. Revenge of the “Sit”: how lifestyle impacts neuro-nal and cognitive health through molecular systems that interface energy metabo-lism with neuronal plasticity. J. Neurosci. Res. 84, 699–715.
Voss, M.W., Nagamatsu, L.S., Liu-Ambrose, T., Kramer, A.F., 2011. Exercise, brain, andcognition across the lifespan. J. Appl. Physiol. 111, 1505–1513.
Weuve, J., Kang, J.H., Manson, J.E., Breteler, M.M.B., Ware, J.H., Grodstein, F., 2004. Phys-ical activity, including walking, and cognitive function in older women. JAMA 292,1454–1461.
