INDIVIDUAL DIFFERENCES IN COGNITIVE TRAINABILITY
IN THE ELDERLY
Elena Amalie Köstler 10373071
Supervisor: Jessika I. V. Buitenweg Co-Corrector: Dr. Jasper Winkel Universiteit van Amsterdam (UvA) 27-05-2016
TABLE OF CONTENTS I. ABSTRACT 3 II. INTRODUCTION 4 III. METHOD 7 A. Participants 7 B. Materials 8 C. Procedure 11 D. Statistical Analysis 13 IV. RESULTS 14 E. Inhibition (Simon) 16 F. Updating (Corsi) 17 G. Switching (TMT) 17 V. DISCUSSION 20 VI. APPENDIX 23 VII. LITERATURE 25
I. ABSTRACT
In 2050 there will be 1.5 billion elderly people. With them there will be a great need to slow down the process of cognitive decline. In this study a three week long cognitive training in the form of a language training was given to 13 monolingual participants above the age 65. The impact of individual differences (age, physical activity, intelligence) on cognitive trainability were assessed with three executive functions (inhibition, updating, switching). Results revealed that with increasing age inhibition ability improved and switching ability declined in posttest compared to pretest. With increasing age executive functioning can deteriorate but also improve. This stresses that cognitive decline can be reduced during aging by easy interventions. Detecting further individual differences can optimize these findings in future research.
II. INTRODUCTION
The increase in the world’s elderly population is surprisingly fast. The age group 65 years or older was estimated to be 8 percent of the world’s population (524 million people) in 2010. It is expected that this number will rise to 1.5 billion people by 2050, 16 percent of the world’s population (National Institute of Aging, 2011). Because of this increase, there is a rising need to maintain quality of life. One important way to do so is by slowing down the process of cognitive decline in late adulthood. According to a study by Lawton et al. (1999, retrieved from Hering et al., 2014) 60 percent of the age group 70 years or older did not wish to live with any form of cognitive impairment. The study revealed that older adults think of cognitive decline as more threatening to quality of life than pain or functional impairment. This clearly stresses the importance of maintaining a high level of cognitive functioning and individual independence for as long as possible by slowing down the process of cognitive decline.
Aging in the human brain can proceed in different ways: one of them is shrinkage in brain regions or in white matter . There can also be accelerating shrinkage in the 1 2
cerebellum and hippocampus during aging (Raz et al. 2005). Aging is also associated with decreasing white matter integrity in the prefrontal regions (Raji et al. 2015). Nevertheless aging is not only limited to shrinkage or decrease. It was long thought that neuronal plasticity only occurs during early age. It is now clear that neuronal plasticity is still possible even in old age, (i.e. 80 years or older (Carretti et al., 2013)). Lovden et al. (2010) describe aging is a dynamically changing process of gaining and losing, instead of a simple decline. This gives rise to hope that interventions such as learning can contribute to reducing cognitive decline. It was indeed found that brain function and structure can be affected by experience-related changes in social environment, cognitive training or physical activity (Raz & Lindenberger, 2013).
caudate, association cortices, entorhinal cortices (Raz et al., 2005)
1
inferior temporal or prefrontal white matter decrease (Raz et al., 2005)
Earlier studies in Neuropsychology showed how fast the brain can adapt to training for example, one study showed that cognitive training already shows effects in the elderly after 4 weeks of training (Anguera et al.,2013). In addition to learning, bilinguals have many advantages compared to monolinguals throughout life. Bilinguals are better in solving problems of attentional control compared to monolinguals (Bialystok, Craik & Rocco, 2006). Bilingual elderly also make more use of general executive functions to manage two languages compared to monolinguals. Monitoring two languages is a special skill of bilinguals and leads to a greater strengthening of executive processes (Bialystok et al., 2006). Given the sharp decline of executive functions after the age of 60 (Treitz, Heyder & Daum, 2007) there is reasonable hope that elderly monolinguals could benefit from learning a second language and therefore improving executive functions. According to Miyake and Friedman (2012) there are three important executive functions. Inhibition, updating and shifting.
In this study we were therefore particularly interested in the role of individual differences in the cognitive trainability of elderly people. We tested whether learning a second language (cognitive training) as a monolingual has an effect on executive functions (shifting, updating, inhibition) and thus slowing down the process of cognitive decline in the elderly.
One shortcoming of most research is nicely explained by Kanai and Rees (2011). Average data comparison does not reveal any information about individual differences, it ignores a large variation in individual response. They state that individual differences are highly consistent across different tests and may therefore be an indication of differences in brain function. Research most often focuses on comparing averages, without revealing the massive individual differences in performance that are mirrored in the variations in the data of individuals. This is why this study focused on individual differences in cognitive trainability in the elderly. Many recent studies have aimed to determine regional brain change during aging and how this process relates to cognitive changes, but the results have been controversial: Raz et al. (2013) found that differences in brain shrinkage in the elderly can already be detected after a short period of time, but that cognitive activity may have a limited effect. Others revealed that cognitive activity in a six months skill
acquisition training can be linked to changes in the brain (Boyke et al., 2008). However, 3
these studies were unable to integrate the full picture of aging, because there is more to aging than seldom changes in the brain and neuroplasticity.
On the one hand it is important to determine which and how brain regions change during training. On the other hand it is very important to detect which individual differences contribute to maintaining quality of life during late adulthood. Little research can be found that addresses the role of individual differences such as intelligence, quality of sleep or age differences in trainability of the aging human brain. Although the role of physical activity and cognitive functioning is partially understood (Oswald et al., 2006), much is left to be discovered. Thus the primary concern of research should be to determine how different lifestyles and individual factors contribute to cognitive training in order to develop a finely-tuned cognitive training (lifestyle) plan for the elderly to maintain or even improve quality of life for as long as possible. It is crucial to focus on who profits from the training and who does not. Ideally researchers should combine neuroimaging data in order to link individual differences in cognitive abilities and their individual factors with differences in brain connectivity, integrity and structure. Other problems with cognitive training research lie in transfer. The problem is how cognitive training in one domain can be transferred to other important tasks one has to fulfill in daily life. Transfer is the generalization of the trained skill to near or distant skills. Many earlier studies lack focus on transfer. We expect that due to the fact that every elderly person speaks at least one language during their life span, learning a second language is a close transfer compared to abstract cognitive training as it is often used in research.
One very important theory in the context of our study is the cognitive reserve theory (CR, Barulli and Stern, 2013). CR states that differences in cognitive functions/ abilities occur through differences in lifetime intellectual activities and other environmental factors. According to this theory, discrepancies between someone’s level of brain pathology (age-related changes) and the observable functional and or cognitive deficits/impairments lie in CR (Barulli & Stern, 2013). Examples of cognitive reserve are
Grey matter increase is visible in the hippocampus of the left hemisphere and the right nucleus accumbens bilaterally
3
life exposures, occupational/educational and leisure activities or social activities contributing to more successful aging (Suchy et al., 2011) dementia or Alzheimer disease (Valenzuela & Sachdev, 2006; Scarmeas et al., 2001). This theory is therefore crucial in the context of cognitive trainability during aging. Closely linked to the cognitive reserve theory are experience-related changes. Those changes are induced by modification of social environments, physical activity, and cognitive training and affect brain structure and function. The response of the brain to increasing or decreasing physical and or cognitive abilities during aging is called neuroplasticity (Kraft, 2012; Zolyniak et al., 2014). According to Lovden et al (2010) it is important to distinguish between the terms plasticity and (cognitive) flexibility. Plasticity is the capacity for changes in flexibility. A non-English speaker for example who learns English increases their brain's flexibility by changing the brain plasticity through learning the language. The learned language skills are the outcome due to changed plasticity. Cognitive flexibility and its resulting plasticity therefore have a key role.
In line with Bialystok et al. (2006) and Miyake and Friedman (2012) we expected cognitive training in the form of language training to have a positive effect on individual cognitive functioning (cognitive flexibility). Individual differences such as age will 4
negatively predict cognitive functioning (Bishop & Yankner, 2010) whereas intelligence will positively predict cognitive functioning (Deary et al., 2010; Kanai & Rees, 2011). Moreover we expected that physically more active elderly people would perform better in cognitive functioning than physically inactive elderly people (Erickson et al., 2011; Kramer et al., 2008).
III. METHOD
A. Participants
There was a control condition in the study (not included in this analysis) which lead to 43 participants in total. Thirteen participants of the experimental condition (6 males, 7
assessed with the 3 cognitive functions: inhibition, updating, shifting (Miyake & Friedman, 2012).
females; mean age = 70.15 years, SD = 4.24 years) of age 65 or older participated in the study. All participants were Dutch speaking monolinguals. Recruitment focused on the enrollment of healthy monolingual elderly subjects who were in good functional and cognitive status. All thirty participants took part in the same brain training in the form of language training (intervention).
B. Materials Intake
An intake questionnaire was obtained from every participant. It was used in order to collect information about age, gender, whether the participant was raised in another language before the age of 7, whether the participant speaks another language fluently (bilingualism) and whether another Romance language had been spoken for longer than 3 months less than 10 years ago. Further questions in the intake survey asked about the participants’ ability to listen to an audio file, whether the participant had a neuropsychiatric diagnosis, was colorblind, deaf or suffering from bad eyesight. Exclusion followed if one or more of these questions were answered with „yes“.
Shipley Institute of Living Scale Part 2
The Dutch version (Schmand & Smedig, 2000) of the Shipley Institute of Living Scale Part 2 (1940) was used in order to quickly assess intelligence (Sines, 1958). The test was taken online and contained 20 questions in total. The participants were instructed to fill in the test (see Figure 1 a) within a 10 minute time limit. The abstract thinking test contained 20 incomplete questions marked with a space sign (-), which had to be filled in order to complete the row. For example „1 2 3 4 5 -“. Scores on this test could vary from 0 to 20, (correct =1, incorrect = 0).
Language Test
The independent variable was the knowledge level of Italian.The language test was taken twice by all participants before the training (pretest, baseline measurement) and after the
intervention (posttest).The pre-(A) and posttest (B) were counterbalanced: one half of the participants first did AB, the other half BA. Each test consisted of 32 questions (64 in total). Participants were asked to either translate words or sentences from Dutch to Italian or vice versa. Participants translated for example „Jan e Petra sono olandesi“ into Dutch, or „de vader“ into Italian. Knowledge of Italian grammar was also assessed, for example „De meervoudsvorm van ‚italiano‘ “(Plural of „italiano“). Participants could score between 0 and 32 per test (in total 64) (0= incorrect, 1= answered correctly but spelling mistake, 2 = correct). The participants also received a printed diary. The instruction was to answer questions everyday about the past day of studying, three weeks long. The diary contained questions such as „Did you study Italian yesterday for at least 30 minutes?“, „Were you physically active yesterday?“, or „How many hours did you sleep last night“.
Three executive functions
The dependent variable of this study was cognitive trainability and individual differences in trainability. Cognitive trainability was tested in the form of cognitive flexibility, assessed by three cognitive functions: inhibition, updating, switching.
1. Simon Test
Cognitive inhibition was assessed using the Simon test (Simon & Wolf, 1963). Every trial began with a waiting signal, a fixation cross (+) displayed in the center of the screen. Either a red or blue box (Figure 1 b) appeared on the right or left side of the screen-center. The task was to respond to the color of the stimuli and not to the location of the stimuli, with either a left key (left Ctrl) for green boxes, and right key (right Ctrl) for blue boxes. There were eight practice trials at the beginning of the experiment followed by 28 trials in total (14 congruent, 14 incongruent), which were displayed in random order. If the eight practice trials contained a mistake, participants obtained more trials to practice. The stimulus onset started the timing and was ended by the participants response on each trial. There was a 500 ms break between each trial onset. Scores can vary from 0 to 24, correct answers were scored as 1, incorrect as 0.
2. Corsi Block-Tapping Task
The Corsi Block Tapping Task (Kessels et al., 2000) is used to assess visuospatial short-term memory (updating information). The test was assessed on the computer with an increasing number of grey cubes (1x1cm) as stimuli (see Figure 1 c). The instruction was to repeat the sequence in the correct order by clicking on them. The test increased in difficulty as the trials proceeded (from 4 to 6 cubes). Visuospatial short-term memory is measured by increasing the length of the sequences. The test starts with a trial sequence containing two blocks. At least one of the two blocks had to be answered correctly in every trial in order to proceed with the next trial of two sequences with increased length. If both trials of one sequence were answered incorrectly, the test stopped. The final score was the number of correctly repeated trials. Scores ranged from 4 to 6 respectively.
3. Trail Making Test (TMT)
The Trail Making Test (Gaudino, Geisler & Squires, 1995) assesses motor speed and visual attention (switching information). The test was taken on the computer. In part A participants had to quickly draw lines and try to connect 25 consecutive numbers by clicking on them. Part B resembles part A but contained numbers and letters (see Figure 1
d). The second part of the test is harder to complete as it is 56 cm longer than part A, containing more visually interfering stimuli than the first part. The time was measured (in seconds) for how long it took to complete part A and part B.
Individual factors
Individual factors were used to gain insight into whether there are individual differences in cognitive trainability in the elderly. The individual factors consisted of age differences (intake), intelligence (Shipley, 1940; Sines 1958) and individual differences in physical activity (in percent) assessed by the printed version of the daily diary that was filled in by each individual participant. Scores of physical activity were calculated by the number of days of physical activity and time (less/more than 30 minutes of activity per day) resulting in a total percentage of physical activity per participant over the 3 week period.
C. Procedure
The study was conducted in line with the research ethics of the University of Amsterdam. Each participant was tested individually in the same laboratory room of the University of Amsterdam. Each participant filled in an online intake questionnaire (see above) for inclusion/exclusion and general information generation, weeks/days before the experiment started. After participants signed informed consent the procedure of the experiment was explained.
The pretest started with the Shipley Institute of Living Scale Part 2 assessing intelligence, the participants received the instruction to answer 20 questions online in 10 minutes. The order of the questions was the same for every participant. Three cognitive flexibility tests followed for baseline measurement. The three tests were 1.Simon task, 2. Corsi Block Tapping Task, 3.Trail Making Test (TMT). The complete session lasted for about 20 minutes in total.
The first test was the Simon test. Instructions were to react as fast as possible, but to make as few mistakes as possible. Every trial began with a waiting signal. The participant was asked to press the left key (left Ctrl) when they saw a green box, and the
right key (right Ctrl) when blue boxes appeared on the screen. The task was to ignore the location of the boxes and to respond only to the color of the stimuli. There were eight practice trials at the beginning of the experiment followed by 28 trials in total (14 congruent, 14 incongruent) displayed in random order.
The second test was the Corsi Block Tapping Test. The participants were instructed to memorize the order of the presented stimuli as well as possible. Their task was to repeat the trials quickly and accurately. The tapping sequence consisted of two demonstration (trials) per sequence. Each sequence increased in length and therefore in complexity. The experiment stopped as soon as the participant repeated a full sequence (2 trials) incorrectly.
The Trail Making Test (TMT) was also taken on the computer. The participants had to click as fast as possible on the stimuli (numbers in Part A or numbers and letters in Part B) displayed on the screen. Their instruction was not only to respond as quickly as possible but also to focus on achieving a high score.
The last task of the pretest was to fill in an online Italian test in order to assess the baseline knowledge of Italian of each participant. The language test consisted of 32 questions. Here participants were asked to translate Italian sentences or words into Dutch or vice versa (10 minutes).
After the testing session each participant received a printed version of the language training book. The instruction was to study 30 minutes per day, five days per week, three weeks long (7.5 hours in total). The experimenter further explained that the participants had to fill in a printed diary they received from the experimenter.
After the three weeks of training the posttest of each participant was assessed, consisting again of the three neuropsychological tests (TMT, Corsi Block Tapping Test, Simon Test) as in the pretests (20 minutes). The language test (10 minutes) was assessed as in the pretest, but counterbalanced. After the posttest was assessed, participants were thanked for their participation and debriefed.
D. Statistical Analysis
The data was checked for normality (P-P plots) and homogeneity. Because we were interested in individual differences, outliers have not been removed from the dataset and data has not been averaged in order to detect inter-individual differences and variation in responses (Kanai & Rees, 2011). A priori g-power calculation was conducted online for 5
Multiple Regression Analysis, resulting in a sample size of 76 participants (medium effect size = 0.15, power = 08, probability level = 0.05). A manipulation check (with a confidence interval of 95 %) was done with a paired-samples t-test in order to assess differences in the experimental group over time (pre- and posttest). This was done in order to gain insight as to whether there are other unknown factors concurrently affecting the active manipulation besides the known manipulation (language training). Multiple Linear Regression Analysis was conducted in order to see if intelligence, age, physical activity predicted a change in cognitive trainability measured by three executive functions (inhibition, switching, updating). Forced entry was used in the analysis. Multicollinearity was checked with the Variance Inflation Factor (VIF) and Tolerance. Durbin Watson was applied for the assumption of Independent errors.
Free Statistics Calculators; Version 4.0: http://www.danielsoper.com/statcalc/calculator.aspx?id=1
IV. RESULTS
The study recruited 26 participants, with 11 participants dropping out. Participants who reported already knowing Italian or to be fluent in another Romance language dropped out. Others reported that studying Italian in three weeks was either too time consuming for them or they underestimated their ability to study a language in such a short period of time. Other reasons for dropping out were sickness or vacation plans. Two other participants did not study Italian during the four week period. Their data was therefore excluded from analysis. In total there were 11 drop-outs and two participants excluded leaving 13 participants. Outliers were not removed, due to their importance for individual difference detection (Kanai & Rees, 2011).
Table 1 Descriptive Statistics
A manipulation check was done with the data of the remaining 13 participants in a paired-samples t-test. On average, language skills improved in posttest (T1, after training) compared to pretest (T0, before training). Average language skills were better after training (M= 26.15, SE = 2.66) compared to before the language training (M = 16.46, SE = 1.85). This difference was significant t (12) = -3.417, p = 0.005, and represented a large effect size, r=.70. Means, standard deviations and sample size of the differences in
Mean, Standard Deviation and Sample Size
Mean Standard Deviation N
Simon Difference (sec) -49,20 134,11 13 Corsi Difference 0,1538 1,41 13 TMT Difference (ms) 17460,46 52644,51 13 Intelligence 15,31 3,30 13 Age 70,15 4,24 13 Physical Activity 70,72 18,60 13
performance (T1-T0) of the three predictor variables (Intelligence, Age, Physical Activity) and the three outcome variables (Simon, Corsi, TMT) can be found in Table 1. A
multiple linear regression was calculated to predict cognitive trainability/ability in relation to individual differences in intelligence, age and physical activity. For more detailed information, see Table 2. The assumption of Independent errors was tested with Durbin Watson and was not violated in any regression model. The data has been checked for Multicollinearity with Tolerance and the Variance Inflation Factor (VIF). This assumption was not violated either. Normality was not violated and checked for by using P-P Plots for each regression model.
Table 2 Linear model of predictors (intelligence, age, physical activity) of cognitive trainability (inhibition, updating, switching).
Multiple Regression Matrix
b β p R2 Simon difference p = .071 0,525 Intelligence 5,67 .14 p = .572 Age -23,647 -.75 p = .017 Physical Activity -2,042 -.28 p = .291 Corsi difference p = .568 0,192 Intelligence -0,114 -.27 p =.410 Age 0,062 .19 p =.588 Physical Activity 0,030 .40 p =.257 TMT difference p = .064 0,537 Intelligence -4716,938 -.30 p =.240 Age 8072,284 .65 p =.030 Physical Activity 397,715 .14 p =.586
E. Inhibition (Simon)
A Multiple Linear Regression Analysis was conducted to see if intelligence, age or physical activity predicted the change in executive functioning over time.
Using the enter method it was found that the individual differences did not significantly explain the amount of variance in the difference on the Simon score (F(3,9) = 3,32, p = .07, R2=.73, R2 Adjusted = .37). The analysis shows that intelligence did not significantly
predict a difference on the Simon score (β = .14, t (12)=.59, p =.57). But age did significantly predict a difference in the Simon score (β = -.75, t (12)=-2.92, p =.017). For more information see Figure 2 and the Appendix. Furthermore physical activity did not significantly predict a difference in the Simon score (β = -.28, t (12)=-1.12, p =.29).
Relationship between Inhibition (Simon) and Age
Simon Dif fer ence in seconds -400 -300 -200 -100 0 100 200 Age in Years 60 65 70 75 80 y = -21,121x + 1432,5 R² = 0,4458
Figure 2: Relationship between Age (in years) and difference score in Inhibition
F. Updating (Corsi)
Furthermore it was found that the individual differences did not significantly explain the amount of variance in the difference on the Corsi score (F(3,9) = 0,71, p = .57, R2=.44, R2 Adjusted = -.08). The analysis shows that intelligence did not significantly predict a
difference on the Corsi score (β = -.27, t (12)=-.86, p =.41). Age did also not significantly predict a difference in the Corsi score (β = .19, t (12)=.56, p =.59). And physical activity did not significantly predict a difference in the Corsi score (β = .40, t (12)=1.21, p =.26). See Table 2 and the Appendix for more information.
G. Switching (TMT)
Individual differences did not significantly explain the amount of variance in the difference on the TMT score (F(3,9) = 3.47, p = .06, R2=.73, R2 Adjusted = .38). The analysis shows that
intelligence did not significantly predict a difference on the TMT score (β = -.30, t (12)=-1.26, p =.24). But age did significantly predict a difference in the TMT score (β = . 65, t (12)=2.57, p =.03), depicted in Figure 3. Additionally physical activity did not significantly predict a difference in the TMT score (β = .14, t (12)=.56, p =.59).
17 Relationship between Switching (TMT) and Age
TMT Dif fer ence in seconds -40 0 40 80 120 160 Age in Years 60 65 70 75 80 y = 8,2777x - 563,25 R² = 0,4443
Figure 3: Relationship between Age (in years) and difference score in Switching (Trail
Data has been plotted in order to compare performance per outcome variable on an individual basis and in order to gain insight into individual change in trainability from pretest to posttest (T1-T0). Figure 4 therefore shows individual difference scores per outcome variable in inhibition (Simon), updating (Corsi) and switching (TMT). We expected that with increasing age cognitive training will have less effect on executive functioning. The first graph shows nicely that inhibition performance improved least in younger participants (aged 65 and 66) and improved most in the oldest participants (aged 76 and 78), which is contradictory to the assumptions. Updating performance (second
graph) is not showing any pattern that could be related to age. Change of performance over time is shown in the third graph and is in line with the assumptions. The older participant group aged 76 improved least in switching over time compared to the younger participants. No participant declined in performance in all three executive functions. Moreover it should be noted that only one participant improved in all three executive functions.
This comparison makes it possible to see in which test and to which extent participants’ performance changed over time. Figure 4 clearly shows that executive functioning did not only change on an inter individual basis but also on an intra individual basis. Cognitive change is therefore much more complex than previously assumed.
V. DISCUSSION
The aim of this study was to determine whether and how strongly individual differences contribute to cognitive trainability in elderly people. Two major findings of this study were that age had an influence on cognitive trainability in different ways. As assumed cognitive training had less impact on switching ability in older participants compared to younger participants. A contradiction to the assumptions was the improvement of inhibition ability with increasing age. Surprisingly, cognitive training had more effect on inhibition performance in older than in younger participants over time. The remaining results of individual differences in intelligence and physical activity were only partly in line with the expectations and were non-significant, leaving two significant findings for further discussion.
The question remains why age had such arbitrary effects on executive functioning. It could be reasoned that switching ability did not change by cognitive training because switching is a more complex cognitive ability than inhibition, switching may need longer and more intensive training in order to improve. Another argument could be that inhibition performance is easily improved by a language training (Kramer & Mota, 2015) and showing strong effects in older adults (Bialystok, Craik,Klein & Viswanathan, 2004). This could explain why inhibition improved after the language training.
The mentioned findings are in line with the cognitive reserve theory (CR, Barulli and Stern, 2013) and experience-related changes theory (Kraft, 2012; Zolyniak et al., 2014), which state that environmental and lifetime changes are crucial for cognitive change. Cognitive change in the ability of inhibition and switching occurred after a three week training, but updating did not significantly change. It could be that a three week long intervention is not long enough to make cognitive change fully possible and/or visible or that the intensity of the intervention was not strong enough. Alternatively individual factors such as age, intelligence or physical activity in combination with cognitive training could have a more complex impact on cognitive functioning than previously expected, which makes it difficult to link individual factors to individual data variations. Beyond that, common statistical analysis with a linear multiple regression could be too mainstream and may not
take individual data variations into account (Kanai & Rees, 2001), thus not allowing inferences on an individual basis to be found within the scope of the study. This study should be repeated with a larger sample size, comparing individual difference data (such as intelligence, age or physical activity) with previous data. This would make it possible to detect patterns of cognitive change and to attribute individual differences to cognitive change/flexibility (Kanai & Rees, 2001).
A number of shortcomings have to be mentioned. Only two of the nine correlations were significant. This could be due to the small sample size (N=13). Non-significant results could possibly have been Non-significant in a bigger sample size. Besides that, recruitment of participants was difficult due to the fact that only healthy monolingual elderly were included who were willing to study Italian for three weeks, leading to a small sample size. A further problem was the high drop-out rate with almost 50% leaving before or during the study. This probably also influenced the data because only highly motivated participants stayed in the study. Another problem of the population of this study was the fact that participants were mainly from the city (Amsterdam) and interested in learning a new language. This are all facts of a homogeneous population which is not representative for the overall population.
Another issue is the fact that every participant claimed to have been physically active almost everyday. This could be due to a misunderstanding of the definition "physical activity" by the participants. Or the homogeneous population is even visible in comparable daily activity scores, leading to the assumption that there was a highly motivated and physically active and vital population recruited for the study.
Variations in task instructions and conductor behavior could have also influenced the performance of the participants. Namely, participants received the task explanations by eight different students, which could have affected the way instructions were interpreted by the participants.
Nonetheless the results show that studying a language for only three weeks already has an impact on two of the three executive functions, from which can be deduced that cognitive change took place. This rises hope for the effectiveness of language training on human cognition. Also major variations in inter- and intra individual differences were
found which are crucial for better understanding how and why people differ from each other and where exactly these differences lie. This is important to know because patterns in cognitive change on an individual basis could help to find the correct intervention at the right time. Before onset of cognitive decline or other age related changes such as dementia. Moreover individual factor detection (intelligence, physical activity, sleep, depression or drug abuse) and their contribution to cognition could help to finetune a cognition training plan. Detecting brain change patterns on individual basis could help to predict future cognitive health conditions, make interventions more effective and lead to a less frustrating and healthier aging style.
On the one hand it became clear that aging is related to decline (in inhibition), on the other hand this study has shown that cognitive change is still possible among elderly people (ability to update). This not only increases hope for this particular field of research but it should also give hope to elderly people who are currently suffering from cognitive decline due to aging. The number of elderly people will constantly rise and with it the great need for help. There is only a limited number of retirement homes. Medical help also knows its limits, which stresses the great need that elderly stay healthy and cognitively active for as long as possible.
VI. APPENDIX
Inhibition (Simon)
The following data refer to Table 2. A non-significant regression equation was found (F (3,9)= 3.317, p = .071), with an
R2 of .525. Participants’ predicted difference in the Simon test (inhibition ability) is equal
to 1667.302 + 5.670 (intelligence) - 23.647 (age) - 2.042 (physical activity), where physical activity is coded in percent. Participants’ ability to inhibit increased 5.670 seconds for each unit of intelligence. But the ability to inhibit decreased by -23.647 for each increased year of age and decreased for -2.042 for each unit of physical activity (in percent). Both intelligence and physical activity were non-significant predictors for the ability to inhibit, whereas age was a significant predictor of inhibition, p = .017. For more information see Table and 2 Figure 2.
Updating (Corsi)
A non-significant regression equation was found (F (3,9)= .714, p = .568), with an
R2 of .192. Participants’ predicted difference in the Corsi test (updating ability) is equal to
-4.592 - .144 (intelligence) + .062 (age) + .030 (physical activity in percent). Participants’ ability to update decreased .114 for each unit of intelligence. Updating increased by .062 for each increased year of age and also increased by .30 for each unit of physical activity (in percent), see Table 2. Intelligence, age and physical activity were
non-significant predictors for updating. Switching (TMT)
A non-significant regression equation was found (F (3,9)= 3.474, p = .064), with an
R2 of .537. Participants’ predicted difference in the TMT test (switching ability) is equal to-
percent). Participants’ ability to switch decreased and therefore improved by 4716.938 milliseconds for each unit of intelligence. The ability to switch increased by 8072.284 for each increased year of age and increased by 397.715 for each unit of physical activity (in percent). Both intelligence and physical activity were non-significant predictors of the ability to switch, whereas age was a significant predictor for switching, p = .030. See table 2 and figure 3 for more information.
VII. LITERATURE
Barulli, D., & Stern, Y. (2013). Efficiency, capacity, compensation, maintenance, plasticity: emerging concepts in cognitive reserve. Trends in cognitive sciences, 17(10), 502-509.
Bialystok, E., Craik, F. I. M., Klein, R., & Viswanathan, M. (2004). Bilingualism, Aging, and Cognitive Control: Evidence From the Simon Task. Psychology and aging, 29, 290- 303.
Bialystok, E., Craik, F. I., & Ruocco, A. C. (2006). Dual-modality monitoring in a
classification task: The effects of bilingualism and aging. The Quarterly Journal of Experimental Psychology, 59(11), 1968-1983.
Bishop, N. A., Lu, T., & Yankner, B. A. (2010). Neural mechanisms of aging and cognitive decline. Nature, 464(7288), 529-535.
Boyke, J., Driemeyer, J., Gaser, C., Büchel, C., & May, A. (2008). Training-induced brain structure changes in the elderly. The Journal of Neuroscience, 28(28), 7031-7035.
Carretti, B., Borella, E., Zavagnin, M., & Beni, R. (2013). Gains in language comprehension relating to working memory training in healthy older adults. International journal of geriatric psychiatry, 28(5), 539-546.
Deary, I. J., Penke, L., & Johnson, W. (2010). The neuroscience of human intelligence dif ferences. Nature Reviews Neuroscience, 11(3), 201-211.
Draganski, B., Gaser, C., Busch, V., Schuierer, G., Bogdahn, U., & May, A. (2004). Neuro plasticity: changes in grey matter induced by training. Nature, 427(6972), 311-312. Erickson, K. I., Voss, M. W., Prakash, R. S., Basak, C., Szabo, A., Chaddock, L., ... & Woj cicki, T. R. (2011). Exercise training increases size of hippocampus and improves memory. Proceedings of the National Academy of Sciences, 108(7), 3017-3022.
Gaudino, E. A., Geisler, M. W., & Squires, N. K. (1995). Construct validity in the Trail Ma king Test: what makes Part B harder?. Journal of clinical and experimental neuro psychology, 17(4), 529-535.
Kessels, R. P. C.; van Zandvoort, M. J. E.; Postma, A.; Kappelle, L. J.; de Haan, E. H. F (2000). "The Corsi Block-Tapping Task: Standardization and Normative Data". App lied Neuropsychology 7 (4): 252–258.
Kok, R. M., & Verhey, F. R. J. (2002). Gestandaardiseerde MMSE. Zeist: Altrecht Ggz
Kraft, E. (2012). Cognitive function, physical activity, and aging: possible biological links and implications for multimodal interventions. Aging, Neuropsychology, and
Cognition, 19(1-2), 248-263.
Kramer, R., & Mota, M. B. (2015). Effects of bilingualism on inhibitory control and working memory: A study with early and late bilinguals. Gragoatá, 20(38).
Kramer, A. F., Erickson, K. I., & McAuley, E. (2008). Effects of physical activity on cognition and brain. Cognitive Neurorehabilitation: Evidence and Applications, 417-434. Lövdén, M., Bäckman, L., Lindenberger, U., Schaefer, S., & Schmiedek, F. (2010). A theo retical framework for the study of adult cognitive plasticity. Psychological bulletin, 136(4), 659.
Miyake, A., & Friedman, N. P. (2012). The nature and organization of individual differences in executive functions four general conclusions. Current directions in psychological science, 21(1), 8-14.
National Institute of Aging, (2011).
Retrieved from https://www.nia.nih.gov/research/publication/global-health-and- aging/humanitys-aging
Oswald, W. D., Gunzelmann, T., Rupprecht, R., & Hagen, B. (2006). Differential effects of single versus combined cognitive and physical training with older adults: the SimA study in a 5-year perspective. European Journal of Aging, 3(4), 179-192.
Raz, N., Lindenberger, U., Rodrigue, K. M., Kennedy, K. M., Head, D., Williamson, A., ... & Acker, J. D. (2005). Regional brain changes in aging healthy adults: general trends, individual differences and modifiers. Cerebral cortex, 15(11), 1676-1689.
Raz, N., & Lindenberger, U. (2013). Life-span plasticity of the brain and cognition: From questions to evidence and back. Neuroscience & Biobehavioral Reviews, 37(9), 2195-2200.
Raz, N., Schmiedek, F., Rodrigue, K. M., Kennedy, K. M., Lindenberger, U., & Lövdén, M. (2013). Differential brain shrinkage over 6months shows limited association with cognitive practice. Brain and cognition, 82(2), 171-180.
Raji, C. A., Eyre, H., Wei, S. H., Bredesen, D. E., Moylan, S., Law, M., ... & Baune, B. T. (2015). Hot Topics in Research: Preventive Neuroradiology in Brain Aging and Co gnitive Decline. American Journal of Neuroradiology, 36(10), 1803-1809.
Scarmeas, N., Levy, G., Tang, M. X., Manly, J., & Stern, Y. (2001). Influence of leisure ac tivity on the incidence of Alzheimer’s disease. Neurology, 57(12), 2236-2242.
Schmand, B. and Smeding, H. M. M (2000). Dutch Translation of the Shipley Institute of Living Scale. Unpublished measure, Academic Medical Center, Amsterdam,
the Netherlands.
Shipley, W. C. (1940). A self-administered scale for measuring intellectual impairment and deterioration. Journal of Psychology, 9, 371-377.
Simon, J. R., and Wolf, J. D. (1963). Choice reaction times as a function of angular stimulus-response correspondence and age. Ergonomics, 6, 99-105.
Sines, L. K. (1958). Intelligence test correlates of Shipley-Hartford performance. Journal of Clinical Psychology.
Suchy, Y., Kraybill, M. L., & Franchow, E. (2011). Instrumental activities of daily living among community-dwelling older adults: Discrepancies between self-report and performance are mediated by cognitive reserve. Journal of Clinical and
Expemental Neuropsychology, 33(1), 92-100.
Treitz, F. H., Heyder, K., & Daum, I. (2007). Differential course of executive control changes during normal aging. Aging, Neuropsychology, and Cognition, 14(4), 370-393.
Valenzuela, M. J., & Sachdev, P. (2006). Brain reserve and dementia: a systematic review. Psychological medicine, 36(04), 441-454.
Zolyniak, N., Schulte-Göcking, H., & Kraft, E. (2014). Neuroplasticity in Aging: Implications for Behavioral and Lifestyle Combined Interventions. Topics in Geriatric
