The Effect of Cognitive and Social Aspects of Musical Activity on the Prevention and Reduction of Age-Related Cognitive Decline

The Effect of Cognitive and Social Aspects of Musical Activity on the Prevention and Reduction of Age-Related Cognitive Decline

Eline L. Bekkers 14-02-2017

Student ID: 10871500

Supervisor: Dr. Hannie Comijs

Co-assessor: Prof. dr. Henkjan Honing

MSc in Brain and Cognitive Sciences, University of Amsterdam Cognitive Science track

Abstract

As the population of older adults will rapidly increase over the coming years, it is becoming increasingly important to discover ways to prevent or reduce declines in cognitive functioning which are observed in healthy ageing. Life-long musical activity seems to be an important protective factor against this age-related cognitive decline. However, little is known about the precise nature of this effect and the precise contribution of different aspects of musical

activity. Therefore, in the current paper, theories on the specific impact of several aspects of musical activity on age-related cognitive decline are discussed. Moreover, their plausibility based on existing literature is evaluated. Convincing evidence exists for beneficial effects of the skill-learning and mental stimulation aspects of musical activity on older adults‟ cognitive functioning through the processes of neuroplasticity and cognitive reserve respectively.

Theories about the specific impact of the multimodal and social aspects of musical activity are supported to a lesser degree, with current research being limited or inexistent. Following the examination of the mechanisms through which musical activity may prevent cognitive decline, the possible modulating effects of several individual differences in musical activity are analysed. Taken together, the discussion in the current paper provides a clearer

understanding of the relationship between musical activity and cognitive decline and

identifies gaps in the current knowledge. As such, it may guide future research on this topic, as well as the use of musical activity as a means to accomplish healthy cognitive ageing.

Introduction

Worldwide, the number of older adults (60 years or over) is growing faster than that of any other age group. By the year 2030, the population of older adults is said to have grown by 56 per cent (United Nations, 2015). Although ageing is widely associated with physical impairments, deterioration of cognitive abilities in many cases causes the need for assistance before a person‟s physical condition does. The care for this ageing population may impact society from a global to individual level, making it essential to discover ways to prevent or reduce cognitive decline.

The rate and severity of age-related cognitive decline is highly variable, with some individuals progressing all the way to a neuropathological state, others merely developing mild difficulties, and yet others hardly developing any overt impairment (e.g., Hayden et al., 2011). This latter group is said to have accomplished „healthy cognitive ageing‟. This steady state of minor decline is widely studied, in the hopes of finding ways to prevent the highly prevalent, more severe forms of cognitive impairment. A random effects model analysis of cognitive decline trajectories in a large cohort of over 1,000 older adults (age 75 and over) showed that, while about two thirds of the cohort was assigned to the healthy ageing group, based on the model, nearly 10% of the cohort was predicted to decline to a level as severe as Alzheimer‟s disease (AD) over the next 12 years (Hayden et al., 2011). Moreover, almost a quarter of the cohort was predicted to decline to a stage in between healthy cognitive ageing and AD, which is referred to as mild cognitive impairment (MCI). In this state, one‟s day-to-day functioning is affected by the loss of cognitive abilities, but not to the extent of a complete loss of independence, which is often the case in AD (Petersen, 2011). MCI should be seen as a serious condition, as it is highly associated with a higher risk of mortality and the development of AD (e.g., Bennett et al., 2002; Palmer et al., 2007) and can affect important daily functioning, such being able to make sound financial decisions and driving (Blazer, Yaffe & Karlawish, 2015). Therefore, it is crucial to discover ways to prevent AD and MCI, and to gain more knowledge on how to accomplish and preserve healthy cognitive functioning in older adults.

Healthy cognitive ageing

Contrary to what the term may suggest, healthy cognitive ageing does not mean a complete absence of cognitive function decline. Small changes in cognitive functioning are an inevitable and universal consequence of structural and functional changes in the brain that occur over the lifetime (Park, O‟Connell & Thompson, 2003). While performance of a small

number of functions improves with age, cognitive functioning in a wide range of domains declines.

Preserved or even improving cognitive functioning is seen for crystallised intelligence, which refers to processes that have been carried out in the past and are dependent on life experience, such as general knowledge and vocabulary (e.g., Christensen, 2001; Salthouse, 2009a). As such, language processes typically remain stable throughout ageing. On the other hand, fluid cognitive processes, which include problem solving, learning and processing new information and manipulation of one‟s environment, peak at around age 30 and then steadily decline with an estimated rate of -0.02 standard deviations per year (Salthouse, 2009b).

Within the attention domain, declines are found in divided attention, as suggested by results from dual-task studies comparing young and older adults (Getzmann, Golob & Wascher, 2016; Verhaeghen & Cerella, 2002). Additionally, older adults‟ ability to ignore irrelevant information while attending to something specific in the environment (i.e., selective attention) seems to be impaired compared to younger adults, with evidence coming from cross-sectional studies using Stroop tasks (e.g., Brink & McDowd, 1999) and visual selection paradigms (e.g., Geerligs, Saliasi, Maurits, Renken & Lorist, 2014; Quigley & Müller, 2014). The impairment of distraction control as observed in the selective attention studies is argued to play an important part in the decline of working memory performance which is often observed in older adults (Darowski, Helder, Zacks, Hasher & Hambrick, 2008). Older adults have a lower working memory span (Bopp & Verhaeghen, 2005), with one cross-sectional study finding that older adults were unable to hold more than four items in working memory, while younger adults could hold five items or more (Schneider-Garces et al., 2010). Episodic memory  information about specific personal life events  also seems to decline gradually throughout adulthood, with an increase in the rate of decline from the age of 60. On the other hand, semantic memory  memory of general world knowledge  increases up to the age of 60 and then declines slightly (Rönnlund, Nyberg, Bäckman, Nilsson, 2005). Interestingly, implicit memory, such as memory for cognitive and motor skills (procedural memory), seems to be preserved (Harada, Love & Triebel, 2013). Memory deficits in older adults seem to result mostly from impaired encoding of new information and retrieval of newly acquired information, while storage of already learned information seems be unaffected (Haaland, Price & Larue, 2003).

Findings on age-related decline of visuo-spatial abilities are inconclusive, as several studies have found age-related declines in tasks involving mental rotation and visuo-spatial imagery (e.g., Berg, Hertzog & Hunt, 1982; Craick & Dirkx, 1992), whereas spatial memory

seems to be unaffected (Parkin, Walter & Hunkin, 1995). De Bruin, Bryant, MacLean & Gonzalez (2016) argue that mostly performance on ecologically relevant spatial tasks is preserved compared to more abstract laboratory tasks.

Older adults perform worse than younger adults on tasks involving aspects of executive functioning, such as attention shifting and learning through trial and error (Grieve, Williams, Paul, Clark & Gordon, 2007), as well as coordination of information streams (Mayr, Kliegl & Krampe, 1996) and inhibition of responses (Wecker, Kramer, Wisniewski, Delis & Kaplan, 2000).

The normal age-related cognitive changes are often said to be due to an overall decline in processing speed (Salthouse, 1996), which in turn causes the impaired performance in the cognitive domains discussed above. On a neural level, ageing is accompanied by a decrease in grey matter, which is the most profound in the prefrontal cortex (Harada et al., 2013). An even greater decrease is seen in white matter (e.g., Meier-Ruge, Ulrich, Brühlmann & Meier, 1992), with some evidence again suggesting the effects are strongest in the prefrontal cortex (Salat et al., 2005).

The decline in the cognitive abilities measured in laboratory tasks (as discussed above) has a very real impact in every-day settings. An interesting observational study, in which performance on „instrumental activities of daily living‟ was compared between young adults (mean age = 22.20) and older adults (mean age = 73.06), showed that healthy cognitive ageing is mostly associated with inefficient performance of actions (e.g., searching multiple areas for objects or having to repeat task steps), causing older adults to take significantly longer to complete an every-day task (Schmitter-Edgecombe & Parsey, 2014). MCI was related to more forgotten and wrong actions, while older adults with AD often lost track of the goal completely and performed irrelevant actions. This shows that, while normal age-related decline still allows older adults to complete day-to-day activities, the effects are noticeable and inconvenient. Therefore, it is not only important to determine how the development of MCI and AD can be avoided, but also how normal age-related impairments can be kept to a minimum.

The degree of cognitive decline in older adults is highly variable (Park et al., 2003), and pathologic indices of dementia only seem to only explain a limited amount of this variation (60%; Boyle et al., 2013). This has inspired researchers to identify predicting factors which are related to late-life cognitive abilities (e.g., Fillit et al., 2002). Several studies have identified lifetime experiences/habits which may prevent or slow the rate of cognitive decline, such as physical activity (Sofi et al., 2011), healthy eating patterns (Hosking, Nettelbeck,

Wilson & Danthiir, 2014) and engagement in social activities (Marioni et al., 2015). For a very detailed discussion of protective factors against cognitive decline and the progression to AD, see Shatenstein & Barberger-Gateau (2015).

Musical activity and cognitive decline

One important possible factor for enhancing or preserving cognitive abilities in older adulthood is the involvement in musical activities throughout the life span. It is well known that music training positively affects cognitive development in early life. For instance, in a longitudinal study, five- to seven-year-old children who had been taking music lessons for a year showed better performance on fine motor skills and auditory discrimination tasks than a non-music control group (Schlaug, Norton, Overy & Winner, 2005). Even more interestingly, though, are the effects of music training on the development of processes outside the auditory-motor domain. For instance, early music training has been linked to improved verbal memory (Ho, Cheung & Chan, 2003), enhanced IQ (Schellenberg, 2004) and increased verbal intelligence and executive functioning in children (Moreno et al., 2011). This shows that musical training may have positive effects on musically related cognitive functions (near-transfer effects) but also more unrelated, higher-order cognitive abilities (far-(near-transfer effects). The fact that the beneficial effects of early music training seem to prevail into adulthood (e.g., Miendlarzewska & Trost, 2014 Shoe & Kraus, 2012) makes it highly likely that musical experiences throughout life have an effect on the decline of cognitive abilities in older age. Supporting this notion, the prevalence of dementia in orchestral musicians has been found to be much lower than what is observed in the general population (Grant & Brody, 2004).

With regards to healthy age-related decline, studies often use cross-sectional designs, comparing cognitive functioning of groups of older musicians and non-musicians. Such studies have revealed that musically trained older adults outperform matched non-musical groups on several musical and non-musical cognitive abilities. A group of older adults who started playing music in early life and had over 10 years of practice (ages 5880) showed better performance on verbal fluency and (working) memory tasks, as well as better visuo-spatial and planning processes (Hanna-Pladdy & Gajewski, 2012). Moreover, a recent ERP study demonstrated that older musicians (mean age = 69.2) made fewer no-go errors on a go/no-go task compared to a matched group of non-musicians, indicating better executive control (Moussard, Bermudez, Alain, Tays & Moreno, 2016).

Although such studies provide more direct evidence for positive link between musical training and cognitive abilities in later life, the results need to be interpreted with some caution, as they cannot with certainty provide information about the causal directionality of the found effects. Several researchers have suggested that differences in brain structure and cognitive functioning between musicians and non-musicians are the result of innate predispositions which happen to differ between the two groups. As such, these pre-existing differences may determine one‟s ability to learn to play a musical instrument and lead them to become a musician (e.g., Zatorre, 2013; Zendel & Alain, 2012).

Another difficulty resulting from cross-sectional studies is the lack of evidence for a differential rate of cognitive decline between musicians and non-musicians, rather than a difference in cognitive abilities at a specific time point. This is the difference between preserved differentiation (persisting enhanced cognitive functioning for musicians versus non-musicians but no differences in the rate of decline) and differential preservation (slower decline of cognitive functioning over time for musicians), as indicated by a main effect of musicianship or an age*musicianship interaction respectively (Zendel & Alain, 2012). A clear schematic illustration of these two concepts is shown in Figure 1.

The problem of innate predispositions is often partially dealt with by careful matching of musician and non-musician groups on several factors, such as age, gender, IQ/education, socio-economic status and physical activity. However, a true causal effect of music training on cognitive functioning and evidence for differential preservation can only be found using a longitudinal study design. The use of such a design could provide evidence for differential rates in are-related cognitive decline between older musicians and non-musicians. However, while longitudinal studies exist on the development of AD and MCI, no such research has been carried out on healthy age-related cognitive decline.

Intervention studies, in which an experimental group of musically naïve older adults receives music lessons for a certain period of time (typically a few months) while a control group does not, do provide some evidence for the effect of music training on cognitive decline. In a study with healthy older adults (ages 6085), 6-month individual piano instruction significantly improved visual processing, planning abilities, and working memory performance compared to an untreated control group (Bugos, Perlstein, McCrae, Brophy & Bedenbaugh, 2007). A less time-consuming approach is the use of statistical models to either predict future decline or to assess the relative contributions of different factors for the decline in cognitive functioning. For instance, Zendel and Alain (2012) measured the performance on several cognitive tasks of a group of musicians and non-musicians with a wide age range

Figure 1. Schematic illustration of the concepts of preserved differentiation and differential

preservation, taken from Alain, Zendel, Hutka and Bidelman (2014).

(ages 1891) and used regression analyses to test for the effects of age and musicianship on cognitive functioning. Differential preservation was found for peripheral and central auditory processing tasks, showing that musical experience can truly affect the rate of cognitive decline throughout adulthood of at least these cognitive functions.

Current study

From these findings, we can conclude that music training throughout the life span may have a significant impact on healthy age-related cognitive decline, preserving not only cognitive abilities directly related to music making, but also affecting cognitive abilities in other domains. However, detailed knowledge on how these beneficial effects come about is lacking. Musical activity is highly complex, placing strong demands on perceptual systems, requiring highly precise motor performance, stimulating complex cortical and sub-cortical processes and influencing relationships with other people. Given this wide variation in different elements involved, it is likely that they may each have distinct effects on cognitive functioning through different underlying mechanisms. Therefore, in this paper, the extent to which four aspects of musical activity may play a unique role in the prevention of cognitive decline are discussed. Specifically, the aspects of skill-learning, multimodal processing, mental stimulation and social aspects are included. By looking into the neural, cognitive and

psychological mechanisms underlying their effects, the workings of the protective effect of musical activity against declines in cognitive functioning in older adulthood can be more thoroughly understood and gaps in the current knowledge may be discovered.

Moreover, unravelling the specific ways in which musical activity may influence older adults‟ cognitive abilities will provide an opportunity to discover factors which may possibly modulate these effects. As the engagement in musical activity varies greatly between people on a number of variables, such as age of onset of musical training, years of experience, frequency of practice, type of music making (different types of instruments) and setting of music making (playing solo versus playing in a group), it is important to determine how these individual differences may influence the protective effects of musical activity against cognitive decline. Examining such effects will help guide future research and may give direction to the practical use of musical activity as a means to accomplish successful cognitive ageing.

In summary, this paper will discuss theories on the distinct effects of several aspects of musical activity on age-related cognitive decline. The extent to which these theories can be supported by the current literature is evaluated and directions for future research are presented. In light of each of the discussed theories, possible modulating effects of several individual factors of musical activity are critically assessed, as they are often ignored, underestimated or overestimated in the current literature.

Skill learning aspect of music training

Playing a musical instrument entails learning a new skill and gaining expertise through extensive practice. Skill learning and expertise are widely studied as modulators of neuroplasticity. Two types of neuroplasticity are typically distinguished: structural and functional neuroplasticity. Structural neuroplasticity refers to changes in grey and white matter volume, which are measured by structural MRI and DTI techniques respectively. Functional neuroplasticity refers to changes in activation of certain brain areas and networks, measured by fMRI techniques. Neuroplasticity has been repeatedly linked to several types of skill learning and expertise (e.g., judo, rock climbing, dance and juggling; see Chang, 2014).

Therefore, this chapter will discuss how the skill learning aspect of musical training can influence cognitive functioning and decline through the mechanism of neuroplasticity. Following this, modulating effects of age of commencement of musical training will be assessed.

Structural neuroplasticity through music practice

A substantial amount of research has focused on the influence of musical training/expertise on structural plasticity, consistently finding differences between adult musicians and non-musicians in grey matter volume in the primary motor cortex. Bangert and Schlaug (2006) found that the effects are dependent on instrument demands, with increased grey matter density in the right precentral gyrus found in adult string players (reflecting the higher demand on the left hand) and a similar effect in the left precentral gyrus in adult piano players (reflecting higher demand on the right hand). The strength of such effects in the precentral gyri seems to be positively related to the degree of musical expertise (Gaser & Schlaug, 2003; James et al., 2014). Similar increases in grey matter volume have been found in auditory and visuo-spatial brain areas. For instance, adult musicians are found to have a greater grey matter density in the superior temporal gyrus (Bermudez & Zatorre, 2005) and Heschl‟s gyrus (Gaser & Schlaug, 2003; James et al., 2014), which are known to be important for pitch processing (Zatorre, Berlin & Penhune, 2002). Increased grey matter density in musicians compared to non-musicians has also been observed in areas involved in visual pattern recognition (right fusiform gyrus) and visuo-motor coordination (inferior temporal gyrus, Gaser & Schlaug, 2003; left intraparietal sulcus, James et al., 2014). Interestingly, increased grey matter density was also discovered in areas involved in processes outside the auditory-motor domain, such as syntactic processing, executive functioning and working

memory (left inferior frontal gyrus and bilateral posterior cerebellar Crus II, James et al., 2014).

Findings on structural differences in white matter tracts between adult musicians and non-musicians are less conclusive. Bengtsson et al. (2005) observed increased fractional anisotropy (FA)  a measure of coherence of aligned fibres  in the internal capsule of the corticospinal tract in musicians compared to non-musicians. Moreover, they found a positive correlation between the estimated hours of music practice during childhood, adolescence and adulthood and FA values in frontal areas and right superior longitudinal fasciculus, which are important for language processing. Similar results were found in a study by Han et al. (2009), with increased FA observed in the corticospinal tract and inferior frontal gyrus. Contrastingly, another study found a decrease in FA values in the corticospinal tract of professional musicians versus that of non-musicians, which  according to the authors  can be due to their larger sample size (Imfeld, Oechslin, Meyer, Loenneker & Jancke, 2009). Lastly, adult musicians seem to have a larger corpus callosum than non-musicians (Schlaug Jäncke, Huang, Staiger & Steinmetz, 1995; Öztürk, Tascioglu. Aktekin, Kurtoglu & Erden, 2002), arguably due to the extensive bimanual motor training that musicians engage in.

Functional neuroplasticity through music practice

Functional differences between musicians and non-musicians are often observed in the primary and secondary motor cortices during the performance of complex finger movements. Decreased activation in these areas has been demonstrated in professional violinists and pianists compared to amateur- or non-musicians during performance of a simple musical sequence (Lotze, Scheler, Tan, Braun & Birbaumer, 2003) or during uni- and bi-manual finger tapping tasks (Jäncke, Shah & Peters, 2000). Similar results have been found in the supplementary and pre-motor areas (Krings et al., 2000). It is hypothesised that these complex finger movements are more automatised in professional musicians, causing them to recruit fewer neurons for the performance of such tasks.

Functional differences in cross-sectional studies are also found in auditory areas, especially in music-related tasks. An interesting study found that musicians have enhanced cortical representation of musical scales in the auditory cortex compared to non-musicians (Pantev et al., 1998). In EEG studies, musicians show an increased mismatch negativity component when detecting melodic deviations (Fujioka, Trainor, Ross, Kakigi & Pantev, 2004).

Interestingly, there is evidence for functional differences outside the auditory-motor domain. An ERP study with chord violation detection demonstrated rapid neural responses (at approximately 200ms) for professional musicians compared to laymen (James, Britz, Vuilleumier, Hauert & Michel, 2008). Source analyses showed the ERPs were generated by right temporal-limbic areas, such as the hippocampal complex, amygdala and right insula, which are associated with cognitive processing, memory and emotion. Moreover, the use of separate working memory systems dedicated to verbal and tonal working memory was observed in musicians compared to a single system used by non-musicians (Schulze, Zysset, Mueller, Friederici & Koelsch, 2011). Furthermore, several studies have demonstrated differences in hippocampal responses between musicians and non-musicians, with Herdener et al. (2010) finding enhanced responses in the hippocampus to acoustic temporal novelty in musicians in both cross-sectional and longitudinal studies. Lastly, a study on semantic memory for melodies revealed higher activity in areas involved in autobiographical memory in musicians versus non-musicians, indicating a constant interaction between episodic and semantic memory (Groussard et al., 2010).

In conclusion, these results from neuroimaging studies demonstrate how music training is positively related to changes in grey matter density, white matter structure, and brain activity in auditory-motor domains, as well as nonrelated domains such as memory, attention and executive functioning.

Music-practice induced neuroplasticity and cognitive decline

The structural and functional changes in the adult musician‟s brain may be an important factor in preventing or reducing the rate of cognitive decline later in life. Older adults‟ susceptibility to functional impairment is said to be dependent on two neural mechanisms, which are related to structural and functional neuroplasticity respectively. Firstly, it depends on structural brain characteristics, such as overall brain volume, grey matter density and white matter integrity (Satz, Cole, Hardy & Rassovsky, 2011). This is referred to as brain reserve capacity (Barulli & Stern, 2013). When this brain reserve capacity is reduced beyond a certain threshold by age-related brain deterioration, functional impairment is observed (Barulli & Stern, 2013). Therefore, enhanced brain reserve capacity acts as s a protection against decline in cognitive functioning.

The threshold for functional impairment is determined by the concepts of neural reserve and neural compensation. Neural reserve refers to the efficiency, capacity and flexibility of innate cognitive networks that have developed over the lifetime. Therefore,

someone with higher neural reserve may show either lower activation (indicating larger network efficiency) or higher activation (indicating larger network capacity) during task performance compared to someone with less neural reserve (Barulli & Stern, 2013). Neural compensation refers to the ability to recruit alternative cognitive networks when important task-related networks are damaged (Barulli & Stern, 2013).

The discussed literature on structural and functional differences between musicians and non-musicians support the notion that musical expertise might influence these neural phenomena. Increased brain reserve capacity through musical training is indicated by the higher grey matter density in auditory-motor and higher-order areas, as well as increased white matter integrity found in the musician‟s brain. Further supporting this, in an MRI study, an absence of significant age-related reductions of grey matter volume and density in the left inferior frontal gyrus and dorsolateral prefrontal cortex was found for musicians compared to non-musicians (Sluming et al., 2002).

Functional differences observed between musicians and non-musicians show that musical training may indeed be linked to higher neural reserve, as the lower activation in motor areas observed in musicians suggests more efficient network recruitment (e.g., Lotze et al., 2003), while the higher activation in auditory areas suggest higher capacity networks for auditory processing of musically-relevant sounds (e.g., Pantev et al., 1998). Lastly, neural compensation in musicians is indicated by their ability to recruit alternate cognitive networks compared to non-musicians, depending on task demands (e.g., different working memory systems in musicians; Schulze et al., 2011).Therefore, the studies on structural and functional brain differences between adults with and without musical expertise support the notion that brain reserve capacity and cognitive network recruitment may be enhanced in the adult brain through structural and functional neuroplasticity caused by music training.

However, it is important to note that the majority of the currently discussed evidence comes from cross-sectional studies. Confounding effects of several individual factors is controlled for in most studies, by matching musician and non-musician groups on at least the factors of gender, age, and education. The alternate explanation of pre-existing differences in brain structure and functioning between musicians and non-musicians cannot be excluded, though. Additionally, the currently discussed studies included participants in mid-adulthood, making it impossible to determine with certainty whether the suggested relationship between musical expertise and enhanced brain reserve and neural network recruitment holds in older adulthood. Although there are studies comparing cognitive abilities of older adults with and without musical training (some of which were discussed in the introduction), evidence for

differences in brain structure and functional network recruitment in older musicians and non-musicians is limited.

Functional differences between older adults with and without musical training are suggested by studies on speech-in-noise perception. For instance, structural equation modelling showed how older adults with previous musical experience maintained their performance on a speech-in-noise perception task by relying more on working memory and attention networks than non-musicians (Anderson, White-Schwoch, Parbery-Clark & Kraus, 2013). This suggests that older musicians‟ auditory processing is boosted through enhanced processing efficiency and recruitment of subtly different cognitive mechanisms compared to non-musicians, supporting the notion that musical expertise is related to enhanced neural reserve and compensation abilities in older adulthood. This in turn may make older musicians less susceptible to functional impairment due to age-related brain deterioration. More such studies comparing older musicians‟ brain structure and activation to that of older non-musicians should be conducted to further endorse this theory.

Modulating effects of differences in onset of music training

The prevention of age-related cognitive decline through neuroplasticity provides more insight into the impact of that the age of commencement of music training may have on this effect. As is known from both animal and human studies, neuroplasticity tends to decrease with age. Therefore, in skill-learning research, the notions of sensitive or even critical periods for learning  a limited or specific time during development when learning or experience can cause long-lasting changes in the brain  are widely discussed (Penhune, 2011).

In the body of literature discussed above, musician adults started music lessons during childhood, nearly always starting before the age of 10, with the average age of commencement around 7 years old. In some cases, age of commencement was not reported, as other criteria for inclusion in the musician group were used (e.g., total years or frequency of music practice). Only a handful of studies have directly tested for effects of age of commencement on the structural or functional differences observed. Firstly, structural differences in the motor cortex seem to depend on age of commencement, as the size of the precentral gyrus of musicians was negatively correlated with age at which music training started (Amunts et al., 1997). Such a relationship between age of commencement and grey matter density is further supported by a recent study, in which early onset music training was related to reduced grey matter volume in the right putamen (Vaquero et al., 2016), compared to late onset training. Moreover, the effect of music training on bi-manual motor performance

may depend greatly on an early start of music training, as indicated by larger corpus callosum sizes only observed in adults who had started music lessons before the age of seven (Schlaug et al., 1995).

This is further supported by results from the DTI study by Bengtsson et al. (2005), in which FA values in the corpus callosum only correlated with musicians‟ estimated hours of practice per week in childhood (< age 11) and adolescence (age 12-16) but not in adulthood (age > 17). These findings are in line with studies suggesting a sensitive period for structural and functional plasticity of the corpus callosum around age six to eight (Westerhausen et al., 2011). Such a dependence on age of commencement for structural changes in white matter were also shown by Imfeld et al. (2009), as diffusivity measures in the corticospinal tract in an early music training group were stronger compared to both a late music training group (based on median split around age 7) and a control group. For functional differences, enhanced cortical representation of music scales was only observed in musicians who started music lessons before the age of nine and was negatively related to age of commencement within this early music training group (Pantev et al., 1998). Such results support the idea of a sensitive period during which music training may affect brain structure and functioning. However, as the early and late commencement groups in these studies both practiced music until the time of testing, an effect of the total years of music training cannot be ruled out (Penhune, 2011). In order to do this, a study should include groups with different total years of training, or else should control for the biological age of the musicians at the moment of testing. The latter was only done in Amunts et al. (1997), with effects holding after controlling for current age of participants, providing strong support for a true effect of age of commencement on brain changes. While Pantev et al. (1998) did not perform such an analysis, it was deemed unlikely that their effects were caused by the duration of training alone, based on the fact that all musicians in the study practiced their instruments for a relatively long time (18 years +/- 5 years). Since both the early and late onset group in Vaquero et al. (2016) had the same level of musical expertise by the time of testing (both groups enrolled in conservatory), the authors speculate that the higher intensity of training for late onset musicians to reach the same level as early onset musicians may have influenced the results, rather than the total years of training.

A second factor that cannot be ruled out in the discussed literature is the frequency of musical practice. Indeed, several studies show that differences in structural or functional plasticity are dependent on the frequency of music practice, rather than the age of commencement or total duration of practice. As mentioned above, grey matter volume in

sensorimotor areas has been found to be enhanced in professional musicians compared to amateurs, with groups only differing in average hours of practice per week or day (Gaser & Schlaug, 2003; James et al., 2014). Similarly, expert pianists performed significantly better on a visual working memory task than amateur musicians, independently of age of commencement of music training (Oechslin, Van De Ville, Lazeyras, Hauert & James, 2013), and had different patterns of brain activation while listening to chord progressions.

Therefore, although the evidence for a sensitive period for the effect of music practice on structural and functional brain characteristics seems consistent, it is possible that the total years of practice or the frequency of practice are confounders in many of the discussed studies. Studies directly comparing the effects of age of commencement, total years of practice and frequency of practice are needed to determine the relative importance of early commencement of music practice.

Lastly, some findings suggest that the sensitive period for music-induced neuroplasticity may not be as strong as it is made out to be. It is known that neuroplasticity through experience is not limited to (early) childhood (e.g., Lövden, Bäckman, Lindenberger, Schaefer and Schmiedek, 2010; Lövden, Wenger, Martensson, Lindenberger & Bäckman, 2013). While structural and functional changes in the brain may be more commonly observed in children than in adults, (due to maturation of brain structures, e.g., Westerhausen et al., 2011), both types of neuroplasticity may still be possible throughout adulthood, and even in older adults (Boyke et al., 2008). In an extensive review, Lövden et al. (2010) discuss how plasticity in adulthood depends on a mismatch between the functioning of existing cognitive networks and task demands imposed by the environment. Such a top-down mechanism may allow older adults to change the structure and efficiency of existing neural networks, or recruit a completely new one, in order to optimally process new bottom-up input. Therefore, learning a new skill in older adulthood may still cause structural or functional changes in the brain, thus resulting in positive effects on cognitive functioning and perhaps the rate of cognitive decline in older adults. Supporting the notion of neuroplasticity in adults, 6-month string instrument training in young adults (ages 2022) with little or no prior experience with this type of instrument, resulted in functional reorganisation of sensorimotor and temporal areas (Kim et al., 2002). In older adults, music lesson intervention showed increased behavioural performance of working memory, visual processing and planning abilities (Bugos et al., 2007). However, neuroimaging studies are needed to prove whether such effects are due to the skill learning (neuroplasticity) aspect of musical training (Lövden et al., 2010).

Summary

There is some evidence that the skill learning aspect of musical training may make older adults with musical expertise less susceptible to functional impairment caused by age-related brain deterioration through the mechanisms of structural and functional neuroplasticity. To further support this notion, more research (preferably longitudinal) into the structural and functional brain differences between musicians and non-musicians in older populations specifically should be conducted. Several findings provide evidence for a sensitive period during which music training can affect structural and functional brain characteristics, with larger effects found for music training started in (early) childhood. This suggests that the age of commencement of musical training may modulate the protective effects of musical activity against cognitive decline. However, direct testing of such effects in older adults is needed. Moreover, the total duration and frequency of music practice have been shown to affect older adults‟ cognitive functioning as well, and cannot be ruled out as confounding factors in most studies comparing early and late commencement groups. Lastly, cognitive improvements following late-life music lessons contradict the existence of a strong sensitive period, although neuroimaging evidence is needed to further prove these effects are dependent on neuroplasticity.

Multimodal aspect of musical activity

Playing a musical instrument places a strong demand on multiple sensory modalities and requires precise motor planning, preparation and execution (Schlaug, Altenmüller & Thaut, 2010). In order to become a proficient player, these sensory and motor systems need to work together efficiently. The sounds produced by the movement output of the motor system are continually assessed by the auditory system, which in turn adjusts the motor output if necessary (Zatorre, Chen & Penhune, 2007). This chapter will examine how the multimodal aspect of music making may influence cognitive functioning and age-related cognitive decline through enhanced auditory-motor coupling. From this analysis, the influence of the variations in the type of musical activity (i.e., playing an instrument, singing or merely music listening) in the prevention/slowing of age-related cognitive decline is discussed.

Enhanced auditory-motor coupling through music training

Music training clearly benefits both auditory and motor functioning independently, as evidenced by enhanced auditory listening skills (e.g., Kishon-Rabin, Amir, Vexler & Zaltz, 2002; Micheyl, Delhommeau, Perrot & Oxenham, 2006) and increased neural efficiency during movement execution in adults (James et al., 2014). However, more interestingly, music training positively affects the way these two systems work together. Several studies have demonstrated activity in auditory areas during motor tasks and activation in motor areas during listening tasks.

This mutual activation is referred to as auditory-motor coupling, and is found to be enhanced in musicians compared to non-musicians. For instance, beginning adult pianists showed additional activity in the motor and premotor cortex when listening to piano melodies (auditory-only task) after 5 weeks of training (Bangert & Altermüller, 2003). The other way around, playing on a muted keyboard (motor-only task) produced additional activity in temporal areas related to auditory processing. Similar results were found in an earlier study, in which pianists showed activity in the primary motor cortex while merely listening to well-practiced piano music (Haueisen & Knösche, 2001). Although a causal conclusion cannot be drawn from these cross-sectional studies, recent findings emerging from a resting state fMRI study strongly suggest that the enhanced functional coupling between auditory and motor systems is indeed caused by musical training. Musicians in this study had increased functional connectivity between the right auditory cortex and ventral premotor cortex compared to non-musicians, even in the absence of an external task (Palomar-Garcia, Zatorre,

Ventura-Campos, Bueicheku & Avilla, 2016). Moreover, the effect was stronger for musicians who started playing earlier in life or had a higher total years of musical training.

Studies on overlapping activity patterns in musicians‟ brains during auditory and motor tasks suggest a musicianship-specific network for auditory-motor integration, consisting of the posterior superior temporal gyrus, the middle temporal gyrus, the supramarginal gyrus and the inferior frontal cortex/posterior frontal operculum (Bangert et al., 2006). The authors argue that that this network may be considered as a „core network‟ for auditory-motor integration in musicians, since activation in these areas is observed in a number of different musical tasks (ranging from listening to performing). A comparable activity pattern was discovered by Schlaug et al. (2005), with musicians having stronger activation in the superior temporal gyrus, posterior inferior frontal gyrus and middle frontal gyrus than non-musicians during music perception tasks.

Music-induced auditory-motor coupling and cognitive decline

The discussed findings show that music training not only benefits the functioning of auditory and motor systems separately, but also positively influences their functional coupling and integration. The unimodal effects of music training on both auditory and motor functioning are important for the study on age-related cognitive decline, since older adults have typically impaired functioning for several aspects of either domain, such as reaction time (Mattay et al., 2002), motor coordination and speed of movement (Kauranen & Vanharanta, 1996), as well as temporal processing (Strouse, Ashmead, Ohde & Grantham, 1998) and central auditory processing (Martin & Jerger, 2005). Interestingly, beneficial effects of music training for unimodal auditory or motor performance are not limited to musically-relevant tasks, but rather carry over to more general auditory and motor abilities. For instance, musicians show fewer age-related declines in central auditory processing tasks (i.e. gap detection) than non-musicians (Alain et al., 2014; Zendel & Alain, 2012). Moreover, older adults with at least one year of music training (mean age = 73.3 years) had fewer errors and faster speed of performance during unimanual and bimanual motor tasks compared to older adults without musical experience (Metzler, Saucier & Metz, 2013).

Unfortunately, there are currently no studies examining the effects of enhanced auditory-motor coupling on the rate of cognitive decline. Studies on its effect on cognitive functioning in general usually focus on improvements on musically-relevant tasks, such as detecting small timing changes (Yee, Holleran & Jones, 1994) and synchronising movements to a beat (Aschersleben, 2002). However, as mentioned by Palomar-Grarcia et al. (2016), it

remains unclear whether the enhanced auditory-motor interaction is related to improved performance on more general, non-musical cognitive tasks. Evidence for such far-transfer effects would show that the multimodal aspect of music training has a positive influence on cognitive abilities in general, and might therefore be a factor in the prevention or reduction of cognitive decline in particular.

One theory suggests that enhanced auditory-motor coupling might play an important role in enhancing/preserving speech processing abilities in older adults, since the increased engagement of fronto-temporo-parietal regions in musicians during both music listening or performance (Bangert et al., 2006) may arguably represent a strengthening (structurally or functionally) of the human mirror-neuron system in musicians compared to non-musicians (Haslinger et al., 2005; Schlaug et al., 2010; Wan & Schlaug, 2010; Zatorre et al., 2007). This system, consisting of the inferior frontal cortex, inferior parietal lobule and superior temporal sulcus (Wan, Demaine, Zipse, Norton & Schlaug, 2010), is believed to be highly important for speech development and perception, following the „motor theory of speech perception‟ (Liberman & Mattingly, 1985). According to this theory, listeners observe the gesticulatory movements of the speaker, which triggers the listeners‟ own gesticulatory motor areas, aiding the interpretation of the speech signal. Evidence for involvement of the motor areas during verbal speech perception tasks supports the notion that an „echo mirror-neuron system‟ may facilitate speech processing, even in the absence of visual information (e.g., Fadiga, Craighero, Buccino & Rizzolatti, 2002; Watkins, Strafella & Paus, 2003; Wilson, Saygin, Sereno & Iacoboni, 2004).

Being able to rely more on the interpretation of activity in gesticulatory motor areas, through strengthening of the mirror-neuron system, may help preserve older musicians‟ speech processing abilities compared to that of non-musicians, allowing them to either compensate for the typically impaired neural encoding of speech (Bidelman, Villafuerte, Moreno & Alain, 2014; Kraus & White-Schwoch, 2014) or general peripheral hearing difficulties. Although there is currently no direct evidence to support this theory, it is interesting to note that several studies have observed enhanced speech perception abilities in older musicians compared to older non-musicians. Positive effects of music training have been found for speech in noise perception (Alain et al., 2014; Parbery-Clark, Strait, Anderson, Hittner & Kraus, 2011; Zendel & Alain, 2012) and phonemic categorisation (Bidelman & Alain, 2015) in older adults. More research into the mechanisms underlying such effects should be conducted in order to determine whether a strengthened mirror-neuron system, due

to enhanced auditory-motor coupling, may contribute to the preservation of these abilities in older musicians.

Taken together, the current literature provides some evidence that the multimodal aspect of musical training may benefit music-related cognitive abilities. Evidence for effects on cognitive abilities outside the musical domain is limited. Following the mirror-neuron system theory, the multimodal aspect of musical activity may preserve the typically declining speech perception abilities in older musicians, although such an effect currently remains speculative.

Modulating effects of different types of musical activity

A possible specific role of the enhanced auditory-motor coupling for the preservation of older adults‟ cognitive functioning implies that the type of musical activity (instrumental or non-instrumental) and perhaps the type of instrument (e.g., keys, strings, wind or vocal) greatly impacts the effect of musical activity on cognitive decline.

As such, several studies provide evidence for the importance of instrumental music practice for the development of long-lasting functional enhancements in children and adults. In a longitudinal study, children who received formal instrumental musical training for 15 months performed better onmusically relevant motor and auditory tasks than a control group which received weekly group music classes consisting of singing and playing with drums and bells (Hyde et al., 2009). In adults, a number of studies directly compared the effects of auditory-only (A) and auditory-sensorimotor music training (AS) on the musically-elicited mismatch negativity ERP component (MMNm) in musically naïve participants. In two studies, the AS group practiced chord progressions on the piano, while the A group listened to the produced chord progressions and judged whether they were correct or not (Lappe, Herholz, Trainor & Pantev, 2008; Pantev, Lappe, Herholz & Trainor, 2009). Before and after the training period, both groups participated in an auditory discrimination test, in which they listened to similar chord progressions (correct and incorrect trials) and judged whether they were correct or not. Meanwhile, MEG data was acquired to determine the strength of the MMNm in both groups. In both studies, the AS group showed increased behavioural performance and significantly stronger MMNm responses to incorrect chord progressions after the training period, whereas the A group did not. A third study using the same design found similar effects for rhythmic piano sequences (Lappe, Trainor, Herholz & Pantev, 2011). These studies show that reorganisation of the auditory cortex, accompanied by enhanced behavioural performance, is only achieved through instrumental music training, rather than

through training involving listening skills only. This supports the notion that instrumental music training in particular may benefit older adults‟ cognitive functioning or prevent age-related cognitive decline through the mechanism of auditory-motor coupling.

This leaves the question whether the type of instrument may modulate the protective effects of musical expertise against cognitive decline. While most studies on co-activation during auditory and motor tasks focused on pianists (Bangert et al., 2006), there is evidence that other types of instrumental training yield similar or at least comparable effects. For instance, professional violinists show increased activation in the right primary auditory cortex areas during the silent performance of a musical piece (no auditory feedback) compared to amateur violinists (Lotze et al., 2003). Furthermore, in a study directly comparing piano players and trumpet players, the latter group showed increased activation in the primary auditory cortex while moving their fingers on a trumpet without auditory feedback, whereas the piano players did not (Gebel, Braun, Kaza, Altermüller & Lotze, 2013). Additionally, while listening to trumpet music (without movement), trumpet players showed increased activation in the primary sensorimotor cortex. Interestingly, whereas the motor activation in piano players is limited to areas representing hand movements, increased activation in trumpet players compared to pianists was observed in areas representing lip movements. These studies show that different types of instrumental training may affect the specific nature but likely not the strength of the resulting auditory-motor coupling, suggesting that differences in the type of instrumental musical activity does not modulate its effect on older adults‟ cognitive functioning.

Interestingly, findings from studies on professional singers suggest that vocal training may have slightly different effects on the connections between the auditory and motor systems. As suggested by Zatorre et al. (2007), the constant monitoring of motor output by the auditory system may be strongest for playing an instrument where pitch is variable and difficult to control, which is the case during vocal performance. Therefore, one may expect the auditory-motor connections to be stronger in singers compared to other instrumentalists. A DTI study provides evidence for this theory, as highly trained singers were compared to a matched group of instrumental musicians and non-musicians (Halwani, Loui, Rüber & Schlaug, 2011). The volume and fractional anisotropy (FA) values of the arcuate fasciculus tract, which connects the superior temporal gyrus, middle temporal gyrus and inferior frontal gyrus, was examined. Note that this network of connected areas is similar to the one identified by Bangert et al. (2006) as the musician-specific network for auditory-motor integration, and greatly overlaps with the mirror-neuron system (see previous paragraph). Although the

arcuate fasciculus had a larger volume and higher FA values in both musician groups compared to non-musicians, singers showed significantly higher arcuate fasciculus volume than instrumental musicians. This suggests that the fronto-temporo-parietal network may be connected more strongly as a result of singing lessons compared to instrumental training.

Summary

The multimodal aspect of musical training might affect musicians‟ cognitive abilities through enhanced auditory-motor coupling, although evidence for improved performance on tasks outside the auditory-motor domain is scarce. Auditory-motor coupling may be specifically linked to the preservation of speech perception capabilities with old age through a strengthened mirror-neuron system. However, such a theory remains speculative at this point. Following the presumed importance of auditory-motor coupling, studies comparing different types of musical engagement support the notion that instrumental music practice is particularly effective for the enhancement of cognitive functioning compared to mere listening practice. Different types of instrumental practice (key, string or wind) seem to be equally effective in accomplishing auditory-motor coupling, although a slight advantage is suggested for vocal training.

Mental stimulation aspect of musical activity

One important reason why musical activity may be so effective in influencing cognitive abilities is its stimulation of not only primary perceptual systems, but also higher-order cognitive processes. Music making requires reading complex, symbolic musical notation and translating this into sequential motor output guided by auditory and somatosensory feedback, as well as memorisation of long musical pieces and possibly even improvisation (Wan & Schlaug, 2010). Music is often compared with language due to its similar reliance on hierarchical syntactic structure (e.g., Rohrmeier, 2011), which features both local and nonlocal dependencies, requiring constant tracking of the short-term and long-term musical context (James et al., 2014). As such, processing the complex hierarchical structure of music greatly involves working memory.

For music listening, this notion is supported by findings from an ERP study, in which music listeners had higher amplitudes of both an early and late negative ERP component in response to a hierarchically irregular versus regular final chord (Koelsch, Rohrmeier, Torrecuso & Jentschke, 2013). Moreover, another study on music listening showed music-induced activation in a network of fronto-temporo-parietal areas, which are involved in top-down attentional and working memory processing (Janata, Tillmann & Bharucha, 2002). With regard to music training, another fMRI study showed that the activity in the fronto-temporo-parietal network was positively related to the level of musical expertise of their participants (Oechslin et al., 2013). Therefore, musical activity stimulates higher-order mental processes (at least attention and working memory), and degree of this stimulation may be dependent on the level of musical expertise.

In this chapter, the relationship between the higher-order mental stimulation aspect of musical activity and older adults‟ cognitive abilities and decline is discussed, as well as the music-specific nature of this effect. Moreover, the impact of continued, life-long musical activity in the prevention of cognitive decline is assessed.

Music-induced mental stimulation and cognitive decline

Stimulation of higher-order cognitive processes is closely related to the cognitive reserve theory of cognitive decline prevention. Cognitive reserve refers to the brain‟s capacity to compensate for damage through the recruitment of pre-existing cognitive resources, therefore maintaining cognitive function at a certain level despite exiting age-related brain deterioration (Stern, 2002). Therefore, cognitive reserve is often used to explain the variance in the functional expression of neurodegenerative brain damage (e.g., Ince, 2001). Research

on risk factors for AD and other dementias suggest that high education, occupation, IQ and participation in leisure activity (physical, mental and social) are protective against functional decline (for review, see Valenzuela & Sachdev, 2006). Therefore, these variables are seen as the indices for cognitive reserve (Jelliger & Attems, 2013). As such, it is believed that consistent engagement of the brain in stimulating activities will preserve cognitive functioning in older adults, following the „use it or lose it‟ principle, which postulates that consistently stimulated cognitive systems are less susceptible to deterioration (Salthouse, 1991).

With regards to music, the cognitive reserve theory predicts that the mental stimulation aspect of music training may preserve older adults‟ cognitive functioning even in the face of age-related brain deterioration. This effect comes about through increased efficiency of existing higher-order cognitive systems and through the ability of musically trained individuals to recruit alternative cognitive networks when tasks become harder to perform. Neuropsychological evidence for this prediction has already been discussed in Chapter 1, in the form of increased neural reserve and neural compensation through functional neuroplasticity in musicians compared to non-musicians. Behavioural studies linking music-induced mental stimulation to higher-order cognitive functioning give more insight into the practical manifestation of enhanced cognitive reserve caused by musical activity.

A large body of literature provides evidence for the prevention of cognitive decline in several domains through participation in mentally stimulating leisure activities in general. In a structural equation modelling study using measurements at three time points over 6 years, intellectually stimulating activities were positively related to changes in cognitive functioning (Hultsch, Hertzog, Small & Dixon, 1999). Moreover, participation in several cognitive leisure activities, including music playing, has been found to delay the onset of rapid memory decline associated with dementia (Hall et al., 2009), to reduce the risk on MCI (Wang et al., 2006), to improve processing speed and memory function (Tesky, Thiel, Banzer & Pantel, 2011) and to slow decline in global cognition, language and executive functioning (Wang et al., 2013). For a recent, very detailed review of more such studies, see Yates, Ziser, Spector and Orrell (2016).

Although playing a musical instrument is often included as a mentally stimulating activity in these studies, the effects of musical activity in particular cannot be assessed from these results. A number of cross-sectional studies do suggest a specific effect of musical activity, by showing how older musicians‟ higher-order cognitive functions are enhanced compared to that of older non-musicians. Hanna-Pladdy and MacKay (2011) showed that

older adults (ages 6083) with over ten years of musical training had better performance on cognitive flexibility, verbal naming and nonverbal memory than older adults with moderate (1-9 years) or no musical training. Similar enhanced performance was found for several verbal abilities (working memory, fluency and general memory), visuo-spatial and planning functions (Hanna-Pladdy & Gajewski, 2012). Moreover, older musicians (ages 5077) outperformed non-musicians on several auditory processing tasks as well as far-transfer tasks, including visuo-spatial span and cognitive control (Amer, Kalender, Hasher, Trehub & Wong, 2013)

Even more importantly, music-induced improvement in higher-order cognitive functioning is supported by results from longitudinal studies. A music intervention study, in which musically naïve older adults (ages 6084) received daily piano lessons for 4 months, showed significant improvement in selective attention and inhibitory control (Stroop task) after the training period compared to a control group (Seinfeld, Figueroa, Ortiz-Gil & Sanchez-Vives, 2013). Additionally, a nearly significant improvement was found for visuo-motor tracking, attention and processing speed (Trail Making Test). Similar results were previously found in Bugos et al. (2007), with improved attentional and working memory functioning (Trail Making Test and Digit Symbol measures) in an experimental piano instruction group compared to a control group. In another longitudinal study, older adults (mean age 77.4 years) with a high level of musical expertise (97.9% could read music and 73.4% were currently musically active) had a higher baseline score on an episodic memory task than a low musical knowledge group (14.3% could read music and 8.1% were currently musically active), and this difference persisted over at least 10 years of annual measurements (Gooding, Abner, Jicha, Kryscio & Schmitt, 2014). Differences in the performance on a semantic memory task did not hold after controlling for general IQ. The ability to read music was positively correlated with over-time performance on both tasks, suggesting that extensive experience with the complex musical notation may be an important factor for stimulating higher-order cognitive systems through music. It would be interesting to find further support for such a special role for reading musical notation.

Taken together, these findings suggest that musical training is indeed related to higher cognitive reserve, in the form of enhanced performance in several higher-order cognitive domains, such as memory, attention and executive functioning. Whether these effects are due to the underlying mechanisms of network efficiency or alternate network recruitment cannot be determined from these behavioural findings.

Effects of musical activity versus other leisure activities

Since several studies have found positive effects on higher-order cognitive functioning by participation in mentally stimulating leisure activities in general, one could argue that the effects of musical training discussed above are not unique to music. However, findings from studies directly comparing the effects of participation in musical and non-musical leisure activities contradict this notion. In the music intervention study by Seinfeld et al. (2013), no improvements in cognitive functioning were observed in the control group, which participated in non-musical leisure activities (i.e., painting, dancing, language lessons, computer lessons and others). Moreover, in the cross-sectional study by Hanna-Pladdy and Gajewski (2012) general activity (i.e., household maintenance, domestic chores, social activities and service to others) did not significantly predict older adults‟ cognitive performance. In a study comparing the effects of older adults‟ participation (ages 7583) in six cognitive and eleven physical leisure activities on their cognitive abilities, a significant association between a higher level of participation in leisure activities and a lower risk of dementia was found (Verghese et al., 2003). Of the physical activities, only dancing had a significant impact on the risk for dementia. For cognitively stimulating activities, doing crossword puzzles, participating in group discussions and writing did not have a significant effect, while playing a music instrument did. However, significant effects of reading and playing board games were also observed, with playing board games having a larger impact that playing a musical instrument. Studies focussing on the effects of other mentally stimulating activities without a direct comparison to musical activity further suggest that music playing may not be unique in its stimulation of higher cognitive function. For instance, older adults‟ engagement in Sudoku or similar puzzles positively correlates with episodic memory and spatial working memory, as well as grammatical reasoning (Ferreira, Owen, Mohan, Corbett & Ballard, 2015). A number of studies finding positive effects of non-musical mentally stimulating leisure activities for the risk of dementia or higher-order cognitive function decline can be found in the review by Yates et al. (2016).

Therefore, it is possible that the mentally stimulating effects of musical activity are not music-specific. However, since the latter mentioned studies do not provide a direct comparison of the effects of musical and non-musical activities, it is impossible to assess the relative strengths of the effects of both types of leisure activity. It would be interesting to investigate in more detail how the effect of music training on the preservation of older adults‟ higher-order cognitive functioning compares to that of other cognitively stimulating leisure activities, in both strength and nature.