The changing brain : neurocognitive development and training of working memory Jolles, D.D.

training of working memory

Jolles, D.D.

Citation

Jolles, D. D. (2011, September 27). The changing brain : neurocognitive development and training of working memory. Retrieved from

https://hdl.handle.net/1887/17874

Version: Not Applicable (or Unknown)

License: Licence agreement concerning inclusion of doctoral thesis in the Institutional Repository of the University of Leiden

Downloaded from: https://hdl.handle.net/1887/17874

Note: To cite this publication please use the final published version (if applicable).

Chapter

In preparation for publication

Dietsje D. Jolles and Eveline A. Crone

Training the developing brain:

a critical evaluation

7

Training the developing brain: a critical evaluation

Abstract

Developmental training studies are important to increase our insight about the po- tential of the developing brain by providing answers to questions such as: “Which functions can and which functions cannot be improved as a result of practice?”,

“Is there a specific period during which training has increased impact?”, and “Is it always advantageous to train a particular function?”. In addition, neuroimag- ing methods may provide valuable information about the underlying mechanisms that drive cognitive plasticity. In this article, we describe how neuroimaging studies of training effects might inform us about the possibilities of the developing brain, and the constraints set by the immature brain structure. At the same time, we also emphasize the complexity of training effects in children. Depending on the type of training and the level of maturation of the individual, training may change de- velopmental trajectories in a different way. It is expected that the immature brain structure might set limits on how much can be achieved with training, but it has been argued that in some cases these same limitations could also be an advantage.

Training the developing brain: a critical evaluation

7.1 Introduction

The human brain is plastic and adapts quickly to new experiences. Several examples are at hand which highlight the plasticity of the brain in adults. For example, the size of brain regions can be modulated by training, which has been demonstrated by the famous studies with London taxi drivers. These studies showed that the size of the hippocampi, which are important for memory, corresponds with the level of driving experience (Maguire et al., 2000; Maguire et al., 2006; see e.g., Elbert et al., 1995; Gaser and Schlaug, 2003 for similar results in musicians). Besides the structure, also the function of the brain is sensitive to experience. This is evident from studies showing altered brain activation in limbic and/or frontoparietal regions for long-term meditation practitioners (Brefczynski-Lewis et al., 2007; Lutz et al., 2008) and after training with working memory (Jolles et al., 2010; Olesen et al., 2004). It is well known that much of our learning takes place in childhood. But what do we know about the plasticity and flexibility of the developing brain? What have neuroscientific studies taught us about training effects in development?

In this article, we will describe a number of theories suggesting that child- hood is a special period during which training and practice may have different ef- fects. However, the article will also show that still relatively little is known about how exactly training-related plasticity differs between children and adults. On the one hand, there are great changes in neural efficiency during development, which could make this period particularly well suited for training interventions. On the other hand, there might also be limitations on the effects of training in childhood.

That is, the maximum achievable performance could be constrained by the current level of structural brain development and cognitive functioning.

Neuroimaging studies may provide extra insights in the underlying cogni- tive and neural processes that are involved during training (cf. Lustig et al., 2009).

In this paper, we particularly focus on neuroimaging studies of training in the do- main of cognitive control and working memory. In adults, these functions are as- sociated with activation in a common set of regions in prefrontal and parietal cortex (e.g., Duncan and Owen, 2000; Owen et al., 2005; Wager and Smith, 2003). Sev- eral studies have reported that activation in these regions decreases after training, particularly when the training led to automatic processing (Chein and Schneider, 2005). On the other hand, when training was specifically directed at the improve- ment of control functions, increases of activation have also been observed (Olesen et al., 2004). There is now a growing interest in the trainability of these regions in children and the difference between training effects in children and adults.

In the following sections, we first give a general introduction about the aims and methods of cognitive training studies. Then, we provide background about the in- teraction between brain maturation and training effects. Finally, we discuss the re- sults of neuroimaging studies of training in adults and we describe the first findings

Training the developing brain: a critical evaluation

in children. We conclude with some critical considerations and directions for future research.

7.2 Cognitive training: purpose and approach

In this article, cognitive training is defined as the process of improving cognitive functioning by means of practice and/or intentional instruction. Before describing neuroimaging findings of developmental training studies and the factors that may contribute to the outcome of training in the developing brain, we briefly outline the goals of training studies and the main methods that are used to achieve those goals.

This short overview is based on both the adult and child literature.

Goal

In general, cognitive training studies have focused on two goals: theory (i.e., getting a better understanding of cognitive plasticity and the underlying neural mecha- nisms), and application (i.e., designing a training intervention which is effective in practice). In this article we mainly focus on the first goal. Therefore, we will not provide an overview of which training methods provide the best results and why, but we will describe how training results are influenced by brain maturation and vice versa. In addition, we describe the possibilities and challenges of using neuroimag- ing methods to examine the underlying mechanisms of training effects.

Paradigm

The major approaches to train cognitive functions can roughly be classified as pro- cess-based or strategy-based training paradigms (cf. Morrison and Chein, 2010;

Noack et al., 2009). First, the process-based approach involves repeated perfor- mance (i.e., practice) of one or more demanding cognitive tasks. Depending on the goals of the study, a variety of different practice paradigms can be used. For example, to answer theoretical questions about particular cognitive processes that are being trained, it is important to keep the training paradigm as simple as possible and control for variables that are extraneous to the trained functions of interest (Luna et al., 2010; Morrison and Chein, 2010). In contrast, to develop training interventions that generalize to real-life situations, it has been suggested that com- plex tasks should be used that train several different processes at the same time (Buschkuehl and Jaeggi, 2010; Green and Bavelier, 2008), and to vary the tasks and stimuli during the training period (Sanders et al., 2002; Schmidt and Bjork, 1992).

Besides, it has been argued that it is important to keep the participant engaged, for example by dynamically adapting the difficulty level (Buschkuehl and Jaeggi, 2010;

Lövdén et al., 2010a).

The second, strategy-based approach uses more explicit instructions that mainly focus on domain-specific processes. For example, to improve the mainte-

Training the developing brain: a critical evaluation nance of information in working memory, it has proven successful to promote re- hearsal, chunking, mental imagery, and/or story-formation strategies (Ford et al., 1984; McNamara and Scott, 2001; St Clair-Thompson et al., 2010; Swanson et al., 2010; Turley-Ames and Whitfield, 2003). In addition, it has proven useful to teach metacognitive strategies about how to approach particular tasks (Ghatala et al., 1985; Kramarski and Mevarech, 2003). For a more detailed description of the dif- ferent types of training, we refer to previous reviews (e.g., Klingberg, 2010; Lustig et al., 2009; Morrison and Chein, 2010; Noack et al., 2009).

Dependent variables

There are several ways to determine the effectiveness of the training, the most ob- vious being performance improvements (e.g., in accuracy or response times) on the trained task. When comparing performance improvements between different groups that are being trained (e.g., children versus adults or children with Atten- tion Deficit Hyperactivity Disorder (ADHD) versus typically developing children), it is important to take into account group differences before and after the training, as well as the room for improvement. For example, if one group performs close to a ceiling level, it is likely that the other group shows larger performance gains. In addition, it is possible that one group shows a larger improvement, while their maxi- mal performance is still below that the other group.

To rule out test-retest effects (Bors and Vigneau, 2001; Goodyear and Douglas, 2009; Jolles et al., 2010), it is important to compare the performance of the trained participants to that of a control group who did not participate in the training. Most studies have used a passive control group, which only participates in the pre- and posttraining sessions. Although a passive control group is useful to rule out the effects of familiarity, it does not take into account expectancy effects and motivation (see Box 2). To control for these effects, an active control group could be included, which receives a placebo treatment (e.g., Klingberg, 2010; Morrison and Chein, 2010). This treatment is difficult to design, because it should be similar to the training program, yet it must not be effective. Therefore, an alternative approach is to compare the effects of two training programs that focus on different cognitive functions (e.g., Mackey et al., 2011).

In addition to improvement on the trained task, several training studies have examined transfer of training effects to untrained tasks. Transfer effects show how training effects generalize to other tasks/functions, and they inform us about the underlying cognitive processes that change as a result of training (e.g., Kling- berg, 2010; Lövdén et al., 2010a). That is, even if one well-described training task is used, there are still many processes that can be influenced by training - an issue that is related to the impurity of executive function tasks (Huizinga et al., 2006;

Miyake et al., 2000). For example, if participants practice with a working memory task, training may lead to a general increase in processing efficiency (e.g., an in- crease of working memory capacity), a strategy change (e.g., the use of rehearsal to

Training the developing brain: a critical evaluation

memorize items in working memory), or a task-specific skill (e.g., familiarity with the memory items). These processes could be disentangled if the participants also perform a number of transfer tasks that have one or more elements in common with the trained task. The use of a latent-variable approach might be particularly fruitful in this respect (Noack et al., 2009; Schmiedek et al., 2010).

Neuroimaging methods provide a promising approach to increase our in- sight in the underlying mechanisms that drive training effects. Several studies have already demonstrated the possibilities of this approach, in which participants are being scanned before and after a period of training (e.g., Dahlin et al., 2008a; Ole- sen et al., 2004; Poldrack and Gabrieli, 2001). The studies performed to date have shown a variety of different effects, including increased activation, decreased activa- tion, or a reorganization of activation. These findings could provide a first step in relating performance changes to the underlying neural mechanisms. One advantage of neuroimaging data is that they can be analyzed along multiple dimensions (e.g., magnitude, location or dynamics of activation and connectivity), which may result in increased sensitivity compared with behavioral measures (Lustig et al., 2009).

Moreover, neuroimaging studies could be used to make predictions about trans- fer effects. For example, Dahlin et al. (2008a) demonstrated transfer effects to an n-back working memory task after 5 weeks of practice with updating in working memory, but there was no evidence of transfer to a Stroop task. Interestingly, it was only the n-back task that showed overlapping striatal activation with the trained task before training, and increases of striatal activation after training. These findings suggest that joint activity in task-specific regions is important for transfer effects to occur.

7.3 Training effects in the context of the developing brain

Developmental cognitive neuroscience is a relatively new field, which has evolved considerably in the last decade. Whereas prior studies were mainly data driven, recently several general frameworks have been proposed to explain the interaction between learning and development. These frameworks are of great importance for understanding training effects in children, given that they provide the intermediate steps for linking brain to behavior.

Theories in developmental cognitive neuroscience

Although the relation between brain maturation and cognitive development has already been emphasized by the classic developmental theories (e.g., Case, 1992;

Piaget and Inhelder, 1974), it is only recently that researchers have been able to provide direct insight into the brain mechanisms underlying cognitive and behav- ioral change. That is, with the appearance of neuroimaging methods, it became pos-

Training the developing brain: a critical evaluation sible to relate neuroanatomical and activation changes to learning and development (cf. Galvan, 2010). According to Johnson (2001, 2011), there are three different viewpoints within the field of developmental cognitive neuroscience. The traditional studies in this field have mostly adopted a maturational viewpoint, which suggests that cognitive functions develop when the underlying brain regions reach maturity.

In other words, children’s immature neural circuitry will constrain the range of possible cognitive processes that could be carried out. However, a growing number of studies now emphasize the role of experience in brain maturation, suggesting that general rules of structural development might be genetically programmed, but specific details are the result of activity-dependent processes influenced by the en- vironment (Changeux and Danchin, 1976; Greenough et al., 1987; Huttenlocher, 2002; Uylings, 2006). These ideas are integrated in a second viewpoint, the interac- tive specialization account, which suggests that the specialization of a particular brain region is a consequence of its interaction and competition with other brain regions over the course of development. This account also points out that brain regions should always be viewed in relation to the functional networks in which they are involved. Finally, the third viewpoint, the skill-learning account, emphasizes that the patterns of change observed during development are sometimes similar to those in- volved during skill acquisition in adults (Casey et al., 2005; Johnson, 2001; Johnson, 2011). This account argues that it is important to distinguish between the effects of age and performance when interpreting developmental differences in brain activa- tion. Together, these viewpoints may be used to describe the effects of training in the developing brain. In the following paragraphs, we describe three questions that are of particular importance when studying the effects of training in children.

Questions regarding training effects in the developing brain

1. How does training influence developmental trajectories?

It has been proposed that there are two ways in which experience can influence the developing brain, that is, via experience-expectant mechanisms and experience-de- pendent mechanisms (Greenough et al., 1987). Experience-expectant mechanisms occur during particular phases of development and are driven by environmental input that is common to all members of a species (such as seeing visual patterns or hearing sounds of speech). On the neural level, these mechanisms might for example involve the overproduction and subsequent pruning of synaptic connec- tions. Experience-dependent mechanisms on the other hand, are driven by input that is specific to an individual. These mechanisms are not strictly age-dependent and involve neural processes that are available throughout lifetime, including the formation of new synapses and changes in the efficiency of synaptic contacts. Ex- perience-expectant and experience-dependent processes probably do not occur independently of one another in driving maturational or learning-related changes (Greenough et al., 1987). It is expected there is always an interaction between the

Training the developing brain: a critical evaluation

two mechanisms.

Recently, Galvan (2010) proposed that development and learning exist on a continuum, with each endpoint receiving inputs from experience-expectant and experience-dependent mechanisms, albeit to a different extent. A graphical presen- tation of this concept is shown in Figure 7.1, which combines the ideas about expe- rience-expectant and experience-dependent changes (Galvan, 2010; Greenough et al., 1987), with the model of life-span cognitive development proposed by Denney (1984; cf. Hertzog et al., 2009). The blue curve represents potential cognitive func- tioning, which increases with development as a result of maturational changes and common environmental experience. However, the increase of cognitive functioning is largest when the individual receives optimal environmental input. Thus, while experience-expectant mechanisms drive the general increase in cognitive function- ing with development, experience-dependent mechanisms determine whether the maximum possible functioning can be reached. Depending on whether the training mainly involves experience-expectant or experience-dependent processes, training may change developmental trajectories in a different way (Figure 7.2). It is impor-

Figure 7.1 This figure represents a simplified (quantitative) representation of how train- ing may affect experience-expectant (EE) and experience-dependent (ED) mecha- nisms (based on Denney, 1984; Galvan, 2010). The blue curve shows the potential of cognitive functioning, which increases with age due to (experience-expectant) matu- ration. In addition, experience-dependent mechanisms determine whether the opti- mally-exercised potential (i.e., the upper limit of cognitive functioning at a certain age;

Denney, 1984) will be reached. It is expected that training effects are always a combi- nation of both (Galvan, 2010).

Training the developing brain: a critical evaluation

tant to note that the maximum value that can be reached, as well as the slope of the curve, are different for each cognitive function in each individual. The inter- and intraindividual differences in the characteristics of the curve are determined by a combination of genetic predispositions and prior experience (Denney, 1984;

Hertzog et al., 2009).

2. Do training effects reflect plastic changes of brain structure or flexibility of brain func- tion?

It is important to note that training-related performance improvements are not necessarily associated with long-lasting plastic changes of the underlying neural cir- cuitry. Performance improvements can also reflect flexibility of brain function that takes place within the limits of the current structural constraints of the brain (Löv- dén et al., 2010a; Posner and Rothbart, 2005). For example, it has been suggested that the failure of young children to rehearse the items that are to be remembered during a working memory task often reflects a production deficiency (e.g., Flavell et al., 1966; Keeney et al., 1967). This indicates that children are able to apply the rehearsal strategy, but they do not always use it. Therefore, training may encourage children to use the strategy, without changing the underlying neural structure (e.g., Ford et al., 1984; Keeney et al., 1967).

Recently, Lövdén et al. (2010a) suggested that structural changes only take place when there is a mismatch between the environmental demands and the pos- Figure 7.2 Depending on the extent to which experience-expectant (EE) and experi- ence-dependent (ED) processes are involved (Galvan, 2010), training might influence developmental trajectories in a different way. In this figure, we present some basic examples of how developmental trajectories might be affected by training. In the first case (graph A), training predominantly influences experience-expectant processes, leading to a speeding-up of development (indicated by a steeper slope), but no dif- ference in the level of cognitive functioning in later in childhood or adulthood. In the second case (graph B), training mainly affects experience-dependent processes, which leads to higher cognitive functioning, but it does not lead to faster development (i.e., higher performance, but no steeper slope). In the third case (graph C), training involves a combination of both processes (i.e., steeper slope and higher performance).

Training the developing brain: a critical evaluation

sibilities of the current structural system. For example, if children practice with a working memory task that requires them to hold more items in mind than they are able to (despite their use of rehearsal strategies), there is a mismatch between the demands of the training paradigm and the supply of the system (i.e., the working memory capacity). As a result, the training might increase working memory capac- ity by inducing plastic changes within the frontoparietal network that is involved in working memory (Klingberg, 2010). The mismatch hypothesis might therefore explain why adaptive training is often more successful than non-adaptive training (Holmes et al., 2009; Klingberg et al., 2005; Klingberg et al., 2002b). Noteworthy, it has been emphasized that a mismatch is a necessary, but not a sufficient con- dition for inducing long-term structural changes (Lövdén et al., 2010a). That is, some structural changes are not possible (e.g., working memory capacity can not be increased infinitely). Moreover, structural changes only occur if training is long enough. Lövdén et al. (2010a) suggested that the minimal length of training that is necessary depends on the type of structural changes that are involved, which could be as fast as a single trial, or require several months to take place. Furthermore, it has been argued that the mismatch should not be too large (i.e., the demands of the environment should not be too high) because participants might give up. Finally, the degree to which plasticity is possible differs between individuals, depending on genetic factors and prior environmental influences.

3. How are training effects influenced by the current stage of development?

We have described that training may influence maturational processes. At the same time, the level of maturation may also influence training effects. Over the course of development, the human brain undergoes dramatic changes, driven by a series of progressive (e.g., myelination and strengthening of synapses) and regressive events (e.g., selective pruning of neurons and synaptic connections) (Giedd and Rapo- port, 2010; Stiles, 2008; Uylings, 2006). It is expected that the same training will have different outcomes in children or adults, depending on the nature of the func- tion which is trained, and the brain structures and neuronal networks in which the changes take place (cf. Galvan, 2010; Kolb et al., 2010). While training in adults mainly modifies the existing neural architecture, in young children it may still in- fluence the construction of neural networks (Galvan, 2010), suggesting that there are both quantitatively and qualitatively different effects of training in children and adults.

On the one hand, the immature brain structure might set limits on how much can be achieved with practice. For example, the speed and efficiency of in- formation processing are determined by the decree of myelination, and the pattern of synaptic connectivity (Chechik et al., 1998; Fields, 2008; Goldman-Rakic, 1987;

Paus, 2010). This could, for instance, constrain practice-related gains on speeded control tasks or working memory (e.g., Case et al., 1982). Besides, training gains are limited by the stage of cognitive development (and thus by age and earlier ex-

Training the developing brain: a critical evaluation perience). That is, a child cannot learn a new skill if this skill builds upon more primitive processes that are not yet mature (Zelazo, 2004). Thus, it is likely that there are cognitive processes that cannot be accelerated with training interventions.

Therefore, it is expected that some age differences are actually magnified rather than reduced after training, which has also been demonstrated in training stud- ies examining younger versus older adults (Baltes and Kliegl, 1992; Nyberg et al., 2003).

On the other hand, it has been suggested that in some cases, immaturity is actually advantageous (Bjorklund et al., 2009; Ramscar and Gitcho, 2007). For example, it has been argued that increasing specialization and integration in brain networks over the course of development goes at the expense of decreasing plastic- ity (Huttenlocher, 2003; Johnson, 2011). Or, as Thompson-Schill et al. (2009) put it: “a system optimized for performance may not be optimal for learning, and vice versa”. In the most extreme case, it has been suggested that there are sensitive periods in brain development during which specific experiences have their largest effects.

Sensitive periods are most pronounced for basic sensory processes that occur dur- ing the first years of life, and they are expected to coincide with periods in which there is an abundance of neurons, axonal projections, and synaptic connections (Greenough et al., 1987; Huttenlocher, 2002; Uylings, 2006). With respect to high- er cognitive functions, there is still a debate about the existence of sensitive periods.

Because of the flexible nature of higher cognitive functions, these functions prob- ably rely on neural mechanisms with life-long plasticity. Nevertheless, it is possible that the capacity for plasticity becomes smaller with age because of the increasing specificity of brain function (Huttenlocher, 2003; Johnson, 2011; Uylings, 2006).

Finally, without denying the possible influence of time-locked biological processes, it is important to note that even (the onset and duration of) sensitive periods are largely influenced by experience. For example, it has been demonstrated that once a neural network is shaped by a particular environmental input, it is dif- ficult to alter the neuronal connections by subsequent experience. These effects are independent of the age of the system (Munakata et al., 2004; Munakata and Pfaffly, 2004). At the same time, if the expected input is not yet received, the network may remain sensitive to new experience for a longer period (Hensch, 2004). Taken to- gether, it seems that the periods of increased sensitivity to training effects are not simply guided by age, but rather by experience-related maturation (Hensch, 2004;

Munakata et al., 2004; Munakata and Pfaffly, 2004).

In summary, we have outlined three issues that are important when interpreting training effects in children: 1. The way training influences developmental trajecto- ries (i.e., does training speed-up development? and does it lead to long-term per- formance benefits?), 2. The cognitive/neural processes that are involved (i.e., does training induce long-lasting plastic changes of the brain structures involved? or does it lead to strategy changes without affecting the underlying brain structure?), and 3.

Training the developing brain: a critical evaluation

The influence of the current level of maturation (i.e., is the immature brain struc- ture a disadvantage or does it lead to greater plasticity?). Next, we will lay out how these issues can be studied using neuroimaging data.

7.4 Neuroimaging studies of cognitive training in adults and children

Immature performance on cognitive control tasks is sometimes mistakenly attrib- uted to age-related maturational constraints. For example, strategy differences be- tween age-matched American and German children suggest that developmental differences in cognitive functioning are sometimes driven by environmental factors, rather than age (Carr et al., 1989). Moreover, several studies have demonstrated that children may improve their performance on cognitive control tasks as a result of training, both in healthy children (e.g., Dowsett and Livesey, 2000; Karbach and Kray, 2009; Thorell et al., 2009), and in children with cognitive or attentional deficits (e.g., Holmes et al., 2009; Klingberg et al., 2005; Klingberg et al., 2002b;

Van der Molen et al., 2010). However, what does it mean if children reach adult levels of performance, or if children with a developmental disability show “normal”

performance after training (cf. Karmiloff-Smith, 2009)? There are a few factors that should be taken into account, including the sensitivity and the ecological validity of the test, and the underlying processes that might be involved. That is, comparable test scores between groups do not necessarily mean that the groups use the same underlying cognitive processes and brain networks. Neuroimaging methods may add to this discussion by giving insight in the underlying mechanisms of cognitive plasticity.

Training the adult brain

To describe the range of possible training outcomes irrespective of the level of mat- uration, we start with an overview of neuroimaging effects of training in adults, with a particular focus on training effects in the domain of cognitive control and working memory. Over the past few years, a growing number of studies have dem- onstrated positive effects of cognitive training in adults (Dahlin et al., 2008b; Jaeggi et al., 2008; Li et al., 2008; Persson and Reuter-Lorenz, 2008; Schmiedek et al., 2010). However, functional magnetic resonance imaging (fMRI) studies of cogni- tive training have shown mixed results. Whereas some studies have mainly reported decreased activation in prefrontal and parietal control regions after practice (Beau- champ et al., 2003; Landau et al., 2004; Qin et al., 2003; Sayala et al., 2006), others have predominantly shown increased activation in these areas (Kirschen et al., 2005; Moore et al., 2006; Nyberg et al., 2003; Olesen et al., 2004). These find- ings suggest that the different tasks/training paradigms are supported by differ-

Training the developing brain: a critical evaluation ent underlying processes. On a cognitive level, there are two general explanations for training-related performance improvements. That is, participants may become more proficient at applying their initial strategy, or they learn to employ a new strat- egy (Jonides, 2004). These types of learning probably involve different underlying neural or cognitive mechanisms, which are summarized below (for an extensive review see Kelly and Garavan, 2005).

Increased proficiency with the initial strategy

When participants show increased competence at applying their initial strategy, they often show a different level of activation within the functional network that they already recruited before practice (Chein and Schneider, 2005; Kelly and Ga- ravan, 2005). In the domain of cognitive control and working memory, the majority of training studies have demonstrated activation decreases in the underlying fron- toparietal regions in this respect, particularly when the training was relatively short.

For example, decreased activation has been observed after training working mem- ory (Garavan et al., 2000; Jansma et al., 2001; Landau et al., 2004; Sayala et al., 2006), visual attention (Tomasi et al., 2004), planning (Beauchamp et al., 2003), and artificial algebra (Qin et al., 2003). There are several possible explanations for these activation decreases, including a shift from controlled to automatic process- ing, increased speed of processing, repetition priming (i.e., implicit memory for task stimuli leading to faster identification), and/or increased specificity of neuronal re- sponses in the underlying neural network (such that firing rate increases for a small set of neurons, while it decreases for the majority of neurons) (e.g., Poldrack, 2000).

It is expected that the magnitude of activation decreases depends on task load. Several studies have suggested that when task load increases, prefrontal activa- tion increases to maintain task performance, but when capacity limits are reached, physiological compensation cannot be made and activation levels off (Callicott et al., 1999; Goldberg et al., 1998; Mattay et al., 2006; Nyberg et al., 2009; Todd and Marois, 2004; Vogel and Machizawa, 2004). Following this line of reasoning, cognitive training should only result in reduced activation if the task is within ca- pacity limits (Nyberg et al., 2009). Consistent with this prediction, frontoparietal activation decreases have been reported after training in working memory updating for young adults, but not for older adults, who likely had a lower working memory capacity (Dahlin et al., 2008a). Moreover, when task load is dynamically adapted to the ability of participants (i.e., by increasing the number of items to be held in working memory), training may even lead to increased frontoparietal activation, as was demonstrated by Olesen et al. (2004). The authors hypothesized that these activation increases reflected an increase of working memory capacity (Klingberg, 2010; Olesen et al., 2004), which is in line with the finding that high performing participants often show higher frontoparietal activation at high task loads than low performing subjects (Gray et al., 2003; Lee et al., 2006; Nyberg et al., 2009).

Finally, several studies of cognitive training have shown increased activa-

Training the developing brain: a critical evaluation

tion in the striatum (Dahlin et al., 2008a; Jolles et al., 2010; Olesen et al., 2004;

Poldrack and Gabrieli, 2001). In line with training studies in the (visuo-) motor domain (Doyon et al., 2009; Hartley et al., 2003; Penhune and Doyon, 2002), these findings may be related to learning-related processes or habit formation (Grahn et al., 2008). On the other hand, the increased striatal activation could also be related to enhancement of task-specific processes (Braver et al., 1997; Lewis et al., 2004;

McNab and Klingberg, 2008; Menon et al., 2000; Postle and D’Esposito, 2003).

Further research is necessary to test between these hypotheses.

The employment of a new strategy

A second possibility for performance improvements is the recruitment of a new strategy. When individuals employ a new strategy, a reorganization of function- al brain networks has often been observed (Kelly and Garavan, 2005; Poldrack, 2000). For example, Poldrack et al. (1998; 2001) showed a spatial shift of activation when participants learned to read a mirror-reversed text. The authors suggested that early in learning, participants mainly relied on basic visual decoding processes, but that they increasingly engaged lexical/phonological processes as learning pro- gressed. Furthermore, it has been suggested that the use of new strategies may lead to increased activation in prefrontal and parietal control regions, even when these strategies lessen task demands (Bor and Owen, 2007b). For example, in a series of experiments Bor et al. (2004; 2003; 2007a) showed that when participants used chunking strategies to maintain information in working memory, frontopari- etal activation increased, although task difficulty decreased. In addition, it has been demonstrated that when participants were trained in using semantic or visuospatial strategies for the encoding of word lists, they showed improved recall and increased activation in frontal and/or occipitoparietal cortex (Miotto et al., 2006; Nyberg et al., 2003). Finally, a strategy change may also induce a shift in the dynamics of activation. For example, using a short strategy training in a group of older adults, Braver et al. (2009) demonstrated a shift from probe-based to cue-based activation in prefrontal cortex regions. This shift was interpreted as a change from a reactive towards a more proactive control mode.

Changes of functional connectivity

In addition to changes in the level of activation within regions, training can also induce changes in the interaction between regions. Such interactions can be stud- ied using functional connectivity (i.e., temporal correlations of blood oxygen level dependent (BOLD) signal fluctuations between brain regions) and effective con- nectivity (i.e., the influence that one region exerts over another) (for a detailed dis- cussion of these concepts, see Friston, 1994). For example, connectivity changes have been observed during artificial grammar learning (Fletcher et al., 1999), rep- etition suppression (Buchel et al., 1999), visual categorization learning (DeGutis and D’Esposito, 2009), and in experts versus non-experts during a creativity task

Training the developing brain: a critical evaluation (Kowatari et al., 2009). Moreover, training-related changes of functional connectiv- ity have been observed during resting-state (Albert et al., 2009; Lewis et al., 2009), suggesting that changes of interregional interactions are not necessarily specific to task conditions.

Functional flexibility and plastic changes in brain structure

Taken together, training may change the level of activation within- and functional connectivity between brain regions involved in the task, depending on the strate- gies that are used and the difficulty of the task. It remains to be determined to which extent these effects reflect flexible changes within the limits of the current physical system, or long-term plastic changes of that system. Long-lasting plastic changes could be associated with a multitude of different structural changes, in- cluding changes in the number or efficacy of synapses, myelination, and changes of hormone or neurotransmitter systems. There are a few neuroimaging studies that have already observed structural changes after cognitive training, including changes in grey- and/or white matter structure (Draganski et al., 2006; Lövdén et al., 2010b;

Takeuchi et al., 2010), and changes in the density of dopamine receptors (McNab et al., 2009). However, it is important to note that only a subset of structural chang- es could be examined using neuroimaging methods, and these methods are a rather indirect measure of the underlying neuronal changes (Poldrack, 2000).

Training the developing brain

There are several possible ways in which cognitive training might affect brain ac- tivation and/or structure in children. These possibilities can be summarized along the lines of the three viewpoints of cognitive development, put forward by Johnson (2001, 2011). It should be noted that these three possibilities are not mutually ex- clusive, and that training effects most likely reflect a combination of factors.

First, in line with the skill learning account, children might show similar effects as those outlined for adults, but then with an easier task. For example, chil- dren could show a similar decrease of activation in prefrontal and parietal control regions when they shift from controlled to automatic processing. In addition, the learning of a particular task-specific strategy might involve similar processes for children and adults. Second, in line with the interactive specialization account, the training outcome in children could be qualitatively different from that in adults.

While training in children might influence the specialization of functional brain net- works, in adults these networks are most likely already specialized. Finally, in line with the maturational account, it is expected that there are limitations on flexibility and plasticity in children, depending on the current level of structural maturation.

Maturational constraints might prevent children from learning a task, or, if children do learn the task, they may rely on compensatory brain regions (Luna, 2004; Scherf et al., 2006).

Training the developing brain: a critical evaluation

Activation changes

There are only a handful of neuroscientific studies that examined activation changes after cognitive training in children. The first set of studies has demonstrated training effects that are similar to the influence of development. For example, in a recent study, Jolles et al. (submitted) showed that age differences in working memory re- lated activation reduced as a result of practice. Previously, it had been demonstrated that 8- to 12-year-old children did not show increased activation for manipulation of information in working memory above and beyond the regions they used for pure maintenance (Crone et al., 2006). However, after 6 weeks of practice, chil- dren showed a similar frontoparietal activation pattern as was seen in adults, argu- ing against the hypothesis that these regions were “inaccessible” due to immature neural circuitry (Jolles et al., submitted). A similar effect has been described for 6-year-old children who participated in attention training. That is, after training, the children showed more mature performance and an adult-like scalp distribution of event-related potentials (ERPs) (Rueda et al., 2005). Another example comes from the domain of language. It was demonstrated that 6-year-old children who prac- ticed for 8 weeks with grapheme-phoneme conversion, demonstrated more mature activation in the left occipitotemporal cortex (Brem et al., 2010).

Notably, these studies also pointed out that there might be limits on the effects of practice in childhood. That is, while 4-year-old children also improved their performance after attention training, they did not show more mature ERPs, as did the 6-year-olds (Rueda et al., 2005). These findings suggest that training of a particular brain function requires a certain stage of cognitive and/or structural brain development. Moreover, the working memory training study showed that training- related activation changes were only present for low working memory loads (Jolles et al., submitted), suggesting that training effects were influenced by the difficulty of the task.

Finally, there are also studies indicating that children and adults process practiced information differently. For example, after practicing for several days with algebra, children showed reduced activation in prefrontal and parietal cortex and increased activation in left putamen (Qin et al., 2004). In contrast, adults who practiced with a similar task only showed reduced prefrontal activation (Qin et al., 2003). It remains to be determined whether these results indicate increased plastic- ity, or whether they are related to immature processing in children (Luna, 2004).

Structural changes

Finally, cognitive training might also lead to changes in the underlying brain struc- ture. For example, adolescent girls who practiced for 3 months with a visuospatial computer game (i.e., tetris), showed increased cortical thickness in superior frontal and temporal areas (Haier et al., 2009). These effects did not overlap with training- related activation changes, suggesting that structural changes do not necessarily result in changes of activation in the same location. In other domains, such as lan-

Training the developing brain: a critical evaluation guage and music, there are some more examples of structural changes after training in children (Hyde et al., 2009; Keller and Just, 2009; Schlaug et al., 2009).

7.5 Critical considerations and future directions

In the present article, we suggested that training effects are better understood in the context of the developing brain, because they emerge from a dynamic interaction between learning and brain maturation (cf. Galvan, 2010). In addition, by provid- ing a small review of the effects of neurocognitive training studies in adults and children, we illustrated how neuroimaging methods could contribute to our under- standing of the underlying cognitive and neural processes that are involved during training. In this paragraph, we point out the issues that warrant further attention.

Neuroimaging as a tool to study training effects

We have described how neuroimaging tools might be valuable in providing additive insights in the underlying cognitive and neural processes that are involved during training. In addition, neuroimaging data may be more sensitive than behavioral measures (cf. Lustig et al., 2009) and they can be used to make predictions about transfer effects (Dahlin et al., 2008a). However, a serious challenge is the complex- ity of the results. There are multiple cognitive and neural mechanisms that can drive changes in activation or brain structure, and these mechanisms might be different for children and adults. Thus, even if developmental and experience-related changes are similar, they are not necessarily caused by the same cognitive or neural process- es (cf. Klingberg, 2006). Moreover, there is a number of confounding factors that further complicate the interpretation of activation changes after practice, including changes in task performance, scanner instability, or reduced anxiety (Box 1). There- fore, it is important to perform theory-driven experiments with well-described tasks and to control for variables that are extraneous to the trained functions of interest (Crone and Ridderinkhof, 2011; Luna et al., 2010; Poldrack, 2000). In addition, human training studies might be conducted in parallel with animal studies and/or with neural network modeling to create hypotheses about the underlying anatomi- cal, histological, and neurochemical processes that are involved during training.

Prior studies have already demonstrated the value of computational modeling in describing how plasticity and learning may differ between children adults (e.g., Elman, 1993; Thomas and Karmiloff-Smith, 2002). In the future, computational modeling should also be combined with neuroimaging methods to create predic- tions about training-related changes in the fMRI signal (Edin et al., 2009; Edin et al., 2007; Macoveanu et al., 2006).

Training the developing brain: a critical evaluation

Box 1 Confounding factors

It seems that there is a multitude of cognitive and neural processes that may underlie the observed training effects, and these processes might differ be- tween children and adults. Moreover, the interpretation of training effects is further complicated by several confounding factors. Here, we briefly summarize the most important confounding factors and some remedies (see also Church et al., 2010; Galvan, 2010; Morrison and Chein, 2010; Poldrack, 2000):

General confounding factors

•

•

•

•

•

Factors specific to neuroimaging

•

•

Raichle et al., 2001).

•

•

•

•

•

Training the developing brain: a critical evaluation

Individual and environmental effects

We pointed out that there are inter- and intraindividual differences in training out- come, depending on an interaction between genetic differences and prior experi- ence. Individual differences might be evident in the ability to learn from training, the rate of learning, and the maximum level of cognitive functioning that can be achieved (e.g., Mercado, 2008). It is expected that differences in training gain are (partly) mediated by differences in brain structure. For example, there are indica- tions that individual differences in brain structure predict performance improve- ments in adults (Erickson et al., 2010; Golestani et al., 2002). Moreover, Shaw et al.

(2006) demonstrated that there are differences between children in the trajectory of cortical development, with more intelligent children showing a prolonged phase of structural brain maturation compared with less intelligent children. These findings indicate that individual differences in training gain might be influenced by the ma- turity of the underlying brain structure, regardless of the child’s age. One important focus for future research involves the characterization of genetic and environmental factors that define individual differences in training gain, and to determine how these factors are related to differences in structural brain maturation.

Another factor that should be considered when examining training gain is the current environment of the individual. For example, it has been argued that children who receive optimal environmental input and training have a large actual- ized genetic potential (Bronfenbrenner and Ceci, 1994), which suggests that extra training will have less additional value. This may explain why cognitive interven- tion programs are often more effective in children from a low socioeconomic back- ground (e.g., Brooks-Gunn et al., 1992). In a similar vein, it has been suggested

Remedies

Some issues are not as problematic as others, i.e., if they influence all condi- tions/groups evenly. In other cases, it is important to gather information about the possible confounding factors and, if possible, control for these factors. Here, we provide some recommendations to explore/control for confounding factors:

•

•

•

•

•

•

Training the developing brain: a critical evaluation

that functions that are frequently practiced in every-day situations, might be more difficult to train than less practiced functions (Denney, 1984). Finally, according to the time displacement hypothesis (e.g., Bavelier et al., 2010), training may even lead to unanticipated negative effects if the activities it displaces are more beneficial than the training itself.

The problem with the adult role model

A final point concerns the use of the adult brain system as a role model to value children’s performance and brain activation (e.g., Bjorklund et al., 2009; Poldrack, 2010). One might question whether this comparison to adults is valid. Functional brain networks in children are not necessarily a simplified, less efficient version of adult brain networks (cf. Power et al., 2010). Although children’s functional networks are organized in a different way, children show operational functional networks with similar network characteristics as adults (Fair et al., 2009; Supekar et al., 2009). Moreover, the immature brain structure might actually have some important evolutionary benefits (e.g., Bjorklund et al., 2009). For example, it has been suggested that language learning is only successful in neural networks with limited cognitive control and working memory abilities (Elman, 1993; Newport, 1990; Ramscar and Gitcho, 2007; Thompson-Schill et al., 2009). Moreover, with advancing levels of expertise and knowledge, individuals usually develop certain routines, which might impair attentiveness and creativity (cf. Hertzog et al., 2009;

Thompson-Schill et al., 2009). These findings have led to the hypothesis that it is not necessarily beneficial (and in some circumstances even disadvantageous) to ac- celerate the development of cognitive-control abilities in children (e.g., Bjorklund et al., 2009). This hypothesis requires further attention in the future.

7.6 Conclusion

Taken together, training studies may provide insight in the possibilities and lim- itations of cognitive functioning over the course of childhood. There are several approaches to study training effects, but there seems to be a trade-off between maximizing the effectiveness of training and maximizing our understanding of the mechanisms underlying training effects. While the effectiveness of training is prob- ably largest when a complex training paradigm is used, understanding the underly- ing mechanisms might require a simpler paradigm. Neuroimaging methods have a great potential to improve our understanding of the interaction between learning and brain development, but there is a number of challenges to overcome.

We described that training effects in the developing brain are driven by a complex interaction between learning, brain development, genetic differences, and prior experience. Depending on the type of training and the level of maturation of

Training the developing brain: a critical evaluation the individual, training may improve the individual’s actualized genetic potential; it may speed-up development; or both. The immature brain structure might set limits on how much can be achieved with training, but it has been argued that these same limitations could also be an advantage. A better understanding of both the limita- tions and the possible advantages is warranted, as it provides the basis for under- standing brain-behavior relations over the course of development.

Although we must be careful when translating scientific research to practi- cal applications (Bruer, 1997; Goswami, 2006), neurocognitive training studies may have potential for application in practice. That is, eventually they might aid in de- signing education programs and interventions for normally developing children or children with developmental disabilities (Carew and Magsamen, 2010; Goswami, 2006; Posner and Rothbart, 2005). For example, to optimize education programs, it is valuable to know more about how children at different ages learn a particular skill, how the underlying neural circuitry supports different kinds of learning, and whether the learning-related changes reflect flexibility in brain function or more permanent changes of the underlying brain structure (Carew and Magsamen, 2010;

Goswami, 2006; Posner and Rothbart, 2005). In addition, knowledge about chil- dren’s abilities to learn might yield insights about specific learning problems, as seen for example in children with dyslexia or ADHD. When the underlying cause of children’s learning difficulties is better understood, it might be possible to target in- tervention to remediate these difficulties (Goswami, 2006). A number of neuroim- aging studies have already started addressing these issues, and they show promising results (Aylward et al., 2003; Hoekzema et al., 2010; Shaywitz et al., 2004; Simos et al., 2002; Temple et al., 2003).