Cognitive reserve and its implications for rehabilitation and Alzheimer’s disease
According to the Cognitive reserve hypothesis, several factors related to mental engagement, such as level of education, type of occupation, leisure activities and social network, appear to affect the risk of developing clinical dementia. The present article provides an overview of the studies that have investigated the effects of mental engagement and cognitive stimulation specifically on dementia of the Alzheimer’s type (AD). Mental training and cognitive stimulation interventions in AD have been shown to be useful in increasing patients’ ability in performing activities of daily living (ADL), allowing them to maintain relative independence. Since cognitive engagement and stimulation are known to modify the brain processes to perform tasks, by recruiting alternative and more efficient networks, this review is especially focused on cognitive rehabilitation in AD patients, which has been shown to improve their global functioning and cognition. This perspective stresses the idea that cognitive reserve is not a fixed factor, but can be continuously modified by life experiences, even when the brain is already affected by neuropathology.
KeywordsCognitive reserveAlzheimer’s diseaseCognitive rehabilitationCognitive stimulation
Alzheimer’s disease (AD) can be defined as a progressive and neurodegenerative disorder, clinically manifested by cognitive deterioration, impairment of activities of daily living, and a variety of neuropsychiatric symptoms and behavioral disturbances (Cummings 2004). It has been demonstrated that some individuals have a better capacity in coping with the disease, showing a discrepancy between the extension of neuropathology and its clinical manifestations (Katzman et al. 1988). It has been hypothesized that such individual differences could be due to some kind of “cerebral reserve” (Stern 2003).
Stern (2003) distinguished roughly between an active and a passive conception of the reserve. According to passive models, often referred to as “brain reserve models”, individual differences in coping with neuropathology derive from structural features (e.g., brain dimension or number of synapses). According to active models or “cognitive reserve models”, the variability in the clinical manifestation of the neuropathology reflects individuals’ different ability to use more efficient and flexible cognitive strategies, which can emerge from different life experiences. Actually, the demarcation between brain reserve and cognitive reserve cannot be so well determined, firstly because differences in cognitive processing should also have a physiologic basis, and secondly because many of the factors associated with increased cognitive reserve, such as cognitively stimulating experiences, appears to have a direct effect on brain structure (Stern 2009).
Adopting a so-called “active” perspective, this review will focus on how engaging in mentally stimulating activities can help in preventing and coping with the clinical manifestations of AD, consistently with the cognitive reserve (CR) hypothesis. Particular emphasis will be placed on research about how to employ mental training and stimulation in AD rehabilitation, integrating them with standard pharmacological interventions.
The article is divided into two main sections. The first section focuses on how different life features related to mental engagement, such as educational level, occupation, leisure activities and social network, could influence the onset and development of AD. The second section focuses on how to exploit mental engagement in AD rehabilitation, aiming at improving specific tasks as well as global function and cognition. Although the distinction between specific domains and global function, which is in accordance with Clare and Woods (2004), may appear blurred from a practical perspective, we decided to adopt this classification for explicative purposes, aware that the two approaches are sometimes superimposable and can often be integrated.
Life experiences, mental engagement, and Alzheimer’s disease
A great number of studies have assessed how lifestyle can be related to the risk of developing dementia. Most research focused on the relation between educational attainment and AD in large-scale populations, evaluating both prevalence (Hill et al. 1993; Kokmen et al. 1993; The Canadian Study of Health and Aging 1994; De Ronchi et al. 1998; Borroni et al. 2008) and incidence (Stern et al. 1994; Fabrigoule et al. 1995; Jacobs et al. 1995; Letenneur et al. 1999; Nielsen et al. 1999; Qiu et al. 2001; Di Carlo et al. 2002; Lindsay et al. 2002) of the disease. The results coming from these studies indicate a dose–effect relation between level of education and AD, since individuals with higher levels of education show a decreased risk of developing the clinical manifestations of the neuropathology.
By measuring regional cerebral blood flow (rCBF), Stern et al. (1992) found an inverse relation between education and cerebral perfusion, since patients with a higher education had the greatest flow reduction in the parieto-temporal cortex, which is a typical pattern of AD neuropathology. This study clearly shows that education cannot provide any type of immunity to the AD process, since the neuropathology occurs independently of education. However, these results also indicate that education might delay the clinical manifestations of AD, since dementia was not detected in the patients with a higher education until functioning was impaired to the same degree as that in patients with a lower education level. Stern et al. (1992) therefore proposed that AD was present in highly educated patients for a longer time, although it was not clinically evident. Consistent results were obtained in two other studies using single photon emission computed tomography (SPECT, Liao et al. 2005) and fluorodeoxyglucose–positron emission tomography (FDG–PET, Perneczky et al. 2006), indicating that education can provide a reserve that compensates for neuropathological deterioration, delaying the onset of its clinical manifestations.
Some studies indicate that not only reaching a high educational level, but also being employed in a highly mentally demanding occupation is related to a decreased risk of developing AD (Evans et al. 1997; Karp et al. 2004; Stern et al. 1994; Qiu et al. 2003). Anttila et al. (2002) showed that having no main lifetime occupation is also associated with an increased risk of AD. Evans et al. (1997) showed that individuals with lower socioeconomic status (SES), computed considering education, occupational prestige and income, had a higher risk of developing the disease. A study by Stern et al. (1994) indicated that the risk of developing AD was stronger in subjects with both low education and low lifetime attainment, suggesting a synergistic effect of education and occupation.
Education may influence the development of AD in several ways. First of all, education may be viewed as reflecting a “brain reserve” (Katzman 1993), such that in subjects with greater education, increased synaptic density in the neocortex would delay the onset of AD, with a subsequent reduction in the prevalence of the pathology. The fact that subjects with a higher education may have more resistance to the progressive deterioration caused by the loss of neuronal function is also supported, counterintuitively, by the greater cortical flow reduction observed in highly educated patients (Stern et al. 1992). Assuming that the deterioration occurs independently of education, educational level may reflect cognitive capacity or stores that allow delaying the clinical manifestations of AD, which do not become evident until the deteriorations reaches a very high degree. As a result, more AD pathology is needed to cause dementia in patients with higher education.
Education could also be considered as a socializing process promoting certain lifelong learning strategies, encouraging the development of decontextualized thinking and enabling the individual to perform more competently on neuropsychological tests (Gilleard 1997). According to Vance and Crowe (2006), education appears to be an important environmental experience that may enhance neural connectivity, as well as the propensity to engage in mentally stimulating activities throughout life. Finally, education may be a surrogate for an intellectual approach to life events (Friedland 1993), which can lead to a lifelong mental stimulation and an enhanced activation of the brain regions involved in memory and in cognitive processing.
Considering that education and occupational complexity can influence the development of AD supports the idea that engaging in mentally stimulating activities can be beneficial for coping with neuropathology. Moreover, this idea is endorsed by animal studies (Diamond 1993; Nithianantharajah and Hannan 2006), which show that complex environments can promote neuronal growth and neuroplasticity, providing a greater resilience to brain damage.
Several articles show that individuals developing AD usually have few hobbies and are less involved in psychosocial activities during their lifetime (Broe et al. 1990; Crowe et al., 2003; Kondo et al. 1994; Friedland et al. 2001; Fritsch et al. 2001; Wilson et al. 2002; Verghese et al. 2003; Scarmeas et al., 2001). Friedland et al. (2001) reported a reduction of the diversity and intensity of intellectual activities during midlife in individuals with AD, compared with healthy subjects. Similarly, Wilson et al. (2002) found that subjects engaged in frequent cognitive activities such as reading, watching television or card playing were less likely to develop AD than subjects with infrequent activities.
Not only the extent of premorbid cognitive experiences, but also their nature, may affect the clinical expression of AD. This is well documented in a study by Wilson et al. (2000), where premorbid reading activity, assessed by considering both reading frequency and availability of reading resources in the home environment, was shown to be positively associated with faster decline on global cognitive and verbal measures, but not on nonverbal measures. These results are consistent with the ones of Stern et al. (1992), showing that more brain degeneration is needed to cause dementia in persons with higher levels of premorbid skill. Once that AD is clinically evident, however, the protective benefits of premorbid experience are assumed to be substantially decreased, and since AD is more advanced in those with higher premorbid skill levels, they decline more rapidly.
Crowe et al. (2003) claimed that one possible way that intellectual/cultural activities could enhance cognitive function in later life might be by building verbal knowledge, which could in turn promote maintenance of memory skills. Since memory problems are central in AD, activities that promote maintenance of memory could be related to clinical prevention or delay of the disease.
The relation between a mentally engaging lifestyle and AD risk could be confounded by education or occupation. For instance, it is known that different active components of CR, as educational attainment and occupational complexity, are dependent on childhood intelligence (especially in modern Western societies), which is in turn affected by biological influences (Whalley et al. 2004). A higher mental ability can not only lead to better scholastic achievements and more cognitively stimulating occupations, but also to engaging in more intellectual leisure pursuits. Education, occupation, and lifestyle could all be markers of innate capacities, which might refer to both genetic background and early developmental factors (Scarmeas and Stern 2003). Different studies have taken this confounding aspect into account. Karp et al. (2004) showed that the association between the incidence of AD and low educational attainment remained significant also when occupation-based socio-economic status was controlled for. Scarmeas et al. (2001) reported that after controlling for education and occupation, the association between leisure activities and AD risk was still present.
It is possible that some subjects may have a reduction of leisure activity as a result of early, still undiagnosed disease. In this case, a lower engagement in mentally stimulating activities would be a manifestation of dementia, rather than a risk factor. Two studies have partially addressed this possibility by excluding from the analyses subjects with cognitive impairment at baseline evaluation (Scarmeas et al. 2001; Wilson et al. 2002), although the activities were recorded only a few years before dementia incidence. In both studies, subjects with high leisure activity showed a decreased risk of developing AD (respectively a 38% and a 33% reduction) compared to subjects with low leisure activity. Further studies should attempt to solve the temporality and causality issues by considering a longer interval between assessment of the activities and AD diagnosis.
One aspect that is strictly related to lifestyle and mental engagement is social network. Married individuals were found to have a longer survival and a decreased risk of developing AD compared to individuals living alone or unmarried (Fratiglioni et al. 2000; Grundman et al. 1996; Håkanson et al. 2009; Helmer et al. 1999; Seidler et al. 2003; Tsolaki et al. 1997; Wang et al. 2002). This is coherent with findings from the animal literature, showing that increased social interactions result in increased synaptic density and cerebral cortex thickness (Diamond 1993).
The CR perspective received criticism by several authors who did not find evidence of the construct in their studies, especially for what concerns education as a protective factor for the clinical manifestations of the disease (Chandra et al. 1998; Cobb et al. 1995; De Ronchi et al. 1998; Beard et al. 1992). This incongruity may be due to methodological issues, as recruiting patients from different kinds of populations with no comparable mean educational levels. For instance, the study of Cobb et al. (1995), which did not show education to be an influent factor for AD risk, included a higher proportion of high school graduates and subjects with medium to high educational attainment. Therefore, there could possibly be a ceiling effect in the results, as well as a threshold of education, below which there is an increased AD risk. This is consistent with the results of De Ronchi et al. (1998), revealing no difference in AD prevalence between less educated and better educated, but also a higher prevalence of AD among subjects that were not educated at all compared to the ones that had at least a few years of schooling. On the other hand, Chandra et al. (1998) performed a community survey in rural India, where two-thirds of the individuals in the cohort were illiterate, showing no association between AD prevalence and literacy. It is important to note that in this study, overall AD prevalence was extremely low due to low life expectancy (fewer subjects living into the age of risk) and shorter survival with duration of disease. Besides, other confounding environmental risks or protective factors (e.g., related to diet or to gene–environment interaction) in this specific population could partially explain why illiteracy did not correlate with AD prevalence. Similarly, no education effect on AD risk was found in other studies on different ethnic groups, like Caribbean Hispanics and African Americans (Tang et al. 2001), and Yoruba (Ogunniyi et al. 2006), while rural living seemed to be the strongest risk factor for AD. These studies indicate that educational attainment is only one of multiple factors that affect AD risk, and in presence of environmental influences associated to poverty and to low socioeconomic status, its influence may be less evident.
The numerous findings of an influence of education, occupation, leisure activity, and social network on the clinical manifestations of AD indicate that cognitive function is dynamic, rather than static (Vance and Crowe 2006). This is coherent with the “use it or lose it” point of view (Hultsch et al. 1999), such that cognitive processes can be modified by exercise, experience, lifestyle, and personality, as well as with the neuroplasticity model of Vance and Crowe (2006). According to this model, neuroplasticity is the potential for morphological changes in the brain to occur with the exposition to stimuli that promote learning or challenge existing neuronal connections to adapt (Vance and Crowe 2006). Subjects that are stimulated by engaging in novel and differentiated activities are also expected to experience cerebral activation, possibly producing an increase in cortex thickness that would result in CR and maintenance of cognitive ability with age. If neuroplasticity declines, the potential to stabilize or increase CR is reduced (Vance and Crowe 2006). In a following article by Vance et al. (Vance et al. 2010), cognitive status is considered to be dependent on factors that can promote either positive or negative neuroplasticity. Positive neuroplasticity refers to beneficial morphological changes in the brain (e.g., strengthening of dendritic connections) that occur with the experience of novel and changing environmental stimulation, which increase CR. Negative neuroplasticity refers to the ability of the brain to atrophy and weaken dendritic connections, decreasing CR, and is due to an unhealthy lifestyle (e.g., poor sleep hygiene, poor nutrition, substance abuse) and to a lack of novel and challenging environmental stimuli. The approach to CR is hence multifactorial: numerous positive and negative factors interact simultaneously, creating a dynamic flow of cognitive functioning. Not taking into account both positive and negative factors may be one of the reasons why some studies were not successful in recruiting CR in AD patients (Chandra et al. 1998; Tang et al. 2001; Ogunniyi et al. 2006). The distinction between positive and negative neuroplasticity is consistent with the description of AD-related cerebral changes made by Teter and Ashford (2002). According to these authors, plasticity in AD may be a process of compensatory but futile modifications in the brain, which can lead to secondary neurodegenerative effects, resulting in a “plasticity burden”. However, plasticity stimulating interventions, which aim to create neural networks that are able to replace lost function, can be clinically effective.
In agreement with the model of Vance et al. (2010), we claim that factors that support neuroplasticity should be targeted for intervention. For this reason, we propose that in order to promote neuroplasticity, and therefore the maintenance of CR, cognitive activity should be exploited in AD rehabilitation.
Cognitive rehabilitation for Alzheimer’s disease
The repeated finding of a relation between AD and different factors associated with cognitive engagement supports the CR hypothesis, here considered as the ability to use more efficient and flexible cognitive strategies that can be trained by means of continuous mental exercise. If it is true that a mentally stimulating lifestyle can modify the paradigms used by the brain to perform a task, then it can also be hypothesized that an intensive mental training or stimulation might provide a greater resilience in facing the neuropathology, possibly recruiting alternate networks. This perspective emphasizes that cognitive reserve is not fixed, and that at any point during life course it can result from a combination of experiences and exposures. In agreement with Vance and Crowe (2006), we believe that increasing CR can allow the maintenance of everyday functioning for longer periods of time in later life.
In opposition to a widespread fatalistic attitude toward dementia, non-pharmacological interventions have been increasingly promoted in the last years, in order to optimize cognition and affect global functioning (De Vreese et al. 2001). In the following paragraphs, by using a distinction adopted by Clare and Woods (2004), we will differentiate between cognitive training of specific functions, such as memory or language, and general enhancement of cognitive and social functioning. We are aware that this distinction is problematic, since some tasks that are apparently focused on a specific function may require more global and general capacities (e.g., tasks that aim to improve orientations skills). Moreover, it is evident that general cognitive and social functioning involves the integration of single functions such as attention, language, and working memory. Nevertheless, we consider this rough distinction useful to describe the different approaches according to their declared aim (training a specific domain vs. improving global functioning).
Special mention will be made of goal oriented cognitive rehabilitation applied to AD patients (Clare et al. 2010), a promising approach which focuses on the identification of personally relevant goals and on the strategies for addressing them.
Cognitive training of specific functions
Cognitive training involves guided practice on a set of standard tasks, which reflect specific cognitive functions. The rationale of these training methods is that regular practice has the potential to improve or maintain functioning in a given domain (Clare and Woods 2004). In order for the patients to benefit from these methods, the effects of practice should be generalized beyond the immediate training context. The ability to extend what has been learned in a specific context to other contexts has been defined as “transfer” (Barnett and Ceci 2002).
The first cognitive training strategies for patients with AD focused mostly on memory recovery. One attempt to treat memory loss with a cognitive approach employed two training strategies derived from information processing models: a didactic training, in which imagery techniques were taught, and a problem-solving training, which presented practical solutions for everyday difficulties caused by memory loss (Zarit et al. 1982). Results showed a relatively small improvement in memory performance, limited to subjects who had received the problem-solving training.
Other studies showed that improvements for recall in AD patients could be gained with some interventions such as repetition and prolonged inspection time (Heun et al. 1997a, b; Kesslak et al. 1997). In another study, where seven patients were trained with an imagery-based mnemonic on face–name retention (Bäckman et al. 1991), no training gains were obtained in any of the patients but one, indicating that such method is far from being generalizable. It can be pointed out that the mnemonic used in this study drew on abilities that are usually severely impaired in dementia, and that a more effective approach could be based on less compromised abilities, such as motor action. Another study that focused on memory retraining, albeit yielding some improvements, showed that changes were not detectable in everyday life and did not influence patients’ functioning (Cahn–Weiner et al. 2003).
One study successfully used procedural memory stimulation for the rehabilitation of AD patients (Zanetti et al. 1994). The main strength of this kind of stimulation is that it relies upon procedural learning, which is part of implicit memory and is relatively well preserved in AD. It has been demonstrated that the ability of AD patients to learn and retain motor and perceptual skills even across a long retention interval is well preserved (Deweer et al. 1994).
Several studies have shown that errorless learning principles can be useful for memory retraining in AD patients (Clare et al. 1999, 2000, 2001, 2002, 2003; Thivierge et al. 2008). Differently from trial and error strategy, errorless learning prevents subjects from making mistakes by encouraging them only to respond when they are sure they are correct, so that they have the opportunity to experience success at every stage of the learning process (Clare et al. 2000). Moreover, the principle behind errorless learning is that when the subject is allowed to make mistakes during the training, such errors interfere with the learning of the correct information. Eliminating mistakes and rehearsing only the correct information can facilitate learning (Anderson et al. 2001). Clare et al. (1999, 2000, 2001) applied errorless learning principles in training AD patients to learn face–name associations, showing that the ability was maintained up to one year (Clare et al. 2001). A very pragmatic application of memory training with errorless learning can be found in the study of Provencher et al. (2008), which employed the technique in coping with topographic disorientation in AD, showing its effectiveness in improving route finding. In this single-case study, one patient was able to learn and retain the procedural component of a short route in a seniors residence. However, the patient’s ability to locate rooms on a map remained impaired, showing that the route learning did not improve general orientation abilities. The fact that the route finding task did not generalize to other destinations indicates that errorless learning may only lead to a very specific route learning, which can certainly be useful, but not extendable to different circumstances (e.g., routine changes, relocation of a patient’s room).
Another memory rehabilitation technique employed with AD patients is spaced retrieval (McKitrick and Camp 1993; Bird and Kinsella 1996; Camp et al. 1996), often referred to as expanding rehearsal, which facilitates new learning by exploiting implicit memory, having the subject practice recall over increasingly longer periods of time. If the retrieval is successful the preceding interval is doubled, otherwise the patient is told the right answer and is asked to repeat it. The interval is then returned to the last interval that was successful for the patient. Hence, spaced retrieval technique is based on an errorless learning approach (Anderson et al. 2001).
Anderson et al. (2001) compared spaced retrieval with a technique based on the use of audiotapes for memory rehabilitation (Tape Therapy, Arkin 1992). Audiotapes included short segments of facts that were relevant for the patient, followed by related questions. The correct answers to the questions were provided after a brief pause, minimizing mistakes. Like spaced retrieval, tape therapy exploits a conjunction of errorless learning and implicit memory. Both spaced retrieval and tape therapy improved the subjects’ memory ability, but spaced retrieval had more rapid effects and a greater carry-over (Anderson et al. 2001).
Several studies attempted to evaluate the possibility to treat both memory and related language deficits with the support of external aids (Bourgeois 1990; Bourgeois 1992; Oriani et al. 2003). Bourgeois (1990) reported that middle-stage AD patients could be trained to use a prosthetic memory aid (a wallet with printed sentences, pictures, and photographs related to the patient’s life) during conversations, and the use of aids could be maintained once training stopped (Bourgeois 1992). However, changes did not seem to influence the subjects’ behavior in everyday conversation, such that partners did not report significant changes over the period of the study. Oriani et al. (2003) reported that the use of an electronic memory aid (EMA) in patients with mild-to-moderate AD improved patients’ prospective memory more than using a written list to recall. Authors underline the fact that the EMA is a memory task itself, since patients have to remember its aim, how to use it, and to carry it about. Therefore, the utility of such device may be limited to patients with a relatively mild cognitive impairment.
Kawashima et al. (2005) reported that after a six-month training program named learning therapy, comprising reading aloud and arithmetic calculation, participants with a diagnosis of AD improved cognitive functions and verbal communication. According to the authors, one possible explanation of this achievement might be the improvement of executive functions during the training.
General enhancement of cognitive and social functioning
Several cognitive training procedures can be criticized for their lack of generalized effects in daily contexts (Zarit et al. 1982; Bourgeois 1990; Bäckman et al. 1991; Bourgeois 1992; Cahn–Weiner et al. 2003), since their influence is mainly restricted to the targeted tasks and does not extend to other cognitive functions. It is however important to point out that not so many studies have evaluated the transfer of a specific skill from one context to another, and therefore further assessments are required.
It has been argued that an efficient functioning in everyday life depends on the flexible use of different strategies adjusted to different situations, so that single functions as memory cannot be considered as muscles that can simply be trained by means of repeated exercise (Ptak et al. 2010). In agreement with this assumption, numerous approaches aim at improving the individual’s general functioning in everyday life, without focusing on domain-specific abilities. The rationale of this choice is that cognitive functions, such as memory or communication, are not used in isolation, but are strongly integrated with other functions (e.g., attention, problem solving, etc.).
One attempt to improve patients’ global functioning is the reality orientation therapy (ROT), which is a psychosocial intervention for persons that manifest memory loss, episodes of confusion, and time–place–person disorientation (Zanetti et al. 1995). ROT uses social interaction and continuous environmental stimulation, which are thought to provide a greater understanding of surrounding, possibly resulting in an improved sense of control and self-esteem (Spector et al. 2001). More specifically, caregivers and professionals are instructed to present basic orienting information during the interactions with the patients, involving them in what is happening around them and reinforcing their interest in the environment (Van der Linden et al. 2003). This therapy has been widely used with AD patients, for instance teaching them to use watches, diaries and daily behavioral itineraries to retrain verbal and behavioral orientations (Hanley and Lusty 1984; Gotestam 1987). Gerber et al. (1991) compared three groups of patients receiving respectively a reality orientation intervention, a social interaction intervention, and no intervention at all, showing that orientation and language improved significantly in the two treatment groups.
Zanetti et al. (1995) evaluated the effects of a long-term group program of formal ROT, administered five days a week for one month to mild-to-moderate AD patients. Each ROT session consisted of an intensive cognitive training during which information such as date, time, current location, and participants’ names was presented. Patients could intervene with the aid of objects such as notes, calendars or clocks. Results showed that ROT decelerated cognitive decline. Other consistent findings indicate that ROT can also be applied successfully in hospital settings (Raggi et al. 2007).
Similarly to reality orientation, cognitive stimulation is a procedure that involves engagement in a range of group activities and discussions, which aim at a general enhancement of cognitive and social function as opposed to focusing on specific abilities (Clare et al. 2004). Moreover, this approach emphasizes positive feedback and motivation (Quayhagen and Quayhagen 2001). Quayhagen and Quayhagen (1989) reported the efficacy of a dyadic cognitive stimulation program, in which patient and caregiver were involved in communication and problem-solving exercises. Cognitive stimulation appeared to be beneficial in numerous studies (Arkin 1992; Quayhagen et al. 1995; Arkin 1999; Arkin 2000; Quayhagen and Quayhagen 2001; Mahendra and Arkin 2003; Spector et al. 2001; Arkin 2007), also showing that involving spousal caregivers in the cognitive stimulation program (Quayhagen and Quayhagen 1989) and utilizing games (Sobel 2001) lead to significant cognitive improvements.
Computer-based training has recently been introduced in the context of cognitive stimulation. In two interactive computer-based individualized cognitive programs, which used personal and biographical material related to the patient’s environment, (e.g., pictures of the patient’s usual shopping route), patients showed fewer errors, performed tasks faster and required less assistance after training (Hofmann et al. 1996; Hofmann et al. 2003). Other studies employed interactive computer-based training programs (Schreiber et al. 1998; Schreiber et al. 1999) and software for neuropsychological training [NPT], (Cipriani et al. 2006) individualized for each patient. Results showed that daily training with NPT improved performances not only at computer-based trained exercises, but also at traditional neuropsychological tests.
Rehabilitation of daily living activities (e.g., brushing teeth, dressing, paying a cheque) through procedural strategies was also shown to be effective, and improvements could be evident even in daily activities that had not been trained, indicating that functional achievements may be independent of the learning context (Farina et al. 2002; Hirono et al. 1997; Zanetti et al. 1997; Zanetti et al. 2001). Another study combined errorless learning and spaced retrieval techniques for retraining patients in performing instrumental activities of daily living (IADL), such as using a voice mail or managing messages from an answering machine (Thivierge et al. 2008). Although not declaring it explicitly, these studies could be assimilated with the goal-oriented cognitive rehabilitation (Clare et al. 2010), a promising approach that has only recently been adopted for the rehabilitation of AD patients. Clare et al. (2010) claim that cognitive rehabilitation should be an individualized approach, which identifies personally relevant goals and devises strategies for addressing them. The goal-oriented cognitive rehabilitation focuses on reducing functional disability and increasing activity and social participation. In their study, Clare et al. (2010) identified for each participant five personally relevant goals in areas relating to self-care, leisure, and productivity. Examples of therapy goals included remembering activities to do in the house, the use of a mobile phone, and maintaining concentration while cooking. The rehabilitation consisted of eight weekly 1-hour individual sessions, which comprised learning practical aids and strategies, techniques for learning new information, attention and concentration exercises, and techniques for stress management. Participants were encouraged to work on their goals by practicing the techniques between sessions. Caregivers were also involved to support between-session practice. Descriptive statistics indicated that cognitive goal-oriented rehabilitation improved patients’ performance, and the improvement was more evident when a caregiver was available for support. The authors point out that the preliminary positive outcomes of this rehabilitation approach were obtained with a very brief intervention, and that longer and more intensive interventions may produce stronger effects in terms of goal attainment. Research should therefore proceed in the direction of goal-oriented trainings, adopting quantitative reliable measures.
Special mention should also be given to music therapy with AD patients, which is reported to improve significantly the quality of life of patients and caregivers (Koger et al. 1999). In fact, patterns of music cognition may differ from abilities in other domains, and caregivers often describe preserved musical functioning in their relatives or clients with AD (Vanstone and Cuddy 2009). Smith–Marchese (1994) reported an amelioration of cognitive deficits in AD patients that participated for six weeks in musical events, which comprised playing simple musical instruments, singing, and body movement. Thompson et al. (2005) reported that exposure to classical music increased cognitive performance in AD patients, leading to better performances on category fluency tasks. Aldridge (1995) underlined that music therapy can enhance socialization and communication skills, leading to a greater cognitive stimulation of the patient.
Several case studies describe AD patients who actively engage in productive musical activities such as playing musical instruments (Beatty et al. 1988, 1994; Fornazzari et al. 2006). Cuddy and Duffin (2005) reported the case of a patient with severe AD who had preserved recognition memory for familiar melodies. Engaging in musical activities, such as listening to music or playing an instrument, can be one way to enhance cognitive reserve, since music training is known to affect cortex morphology and organization (Schlaug 2001; Ohnishi et al. 2001). Moreover, when music faculties are spared in the presence of AD, enjoyment, and relaxation conveyed by the familiarity with music (Vanstone et al. 2009) can be exploited in rehabilitation.
According to the CR theory (Stern 2002, 2003), individual differences in coping with neuropathology can emerge from different life experiences, such as educational attainment, occupation complexity, mentally stimulating leisure activities, and social network. The present review focused on how mental engagement can influence the clinical manifestations of AD and how this can be exploited in rehabilitation.
The analysis of the literature supports the idea that lifestyle can provide a reserve that allows better coping with AD neuropathology. Education and occupation have been associated with the risk of AD in several population studies, and, for equal clinical conditions, the extent of neuropathology in the brain is greater in more educated individuals, meaning that education can provide a reserve that allows coping better with the disease. Together with these findings, it has been shown that different aspects of lifestyle, such as leisure activities and social network, are predictors of a greater resilience to Alzheimer neuropathology.
Considering cognitive reserve as the ability to utilize more efficient and flexible cognitive strategies, which can be trained by means of a continuous mental exercise, it makes sense to consider the possibility of exploiting this capacity in the context of AD rehabilitation. The idea is that CR is not a fixed factor, but is rather continuously modified by environment and life experiences through the whole life course. If it is true that education, occupation, and lifestyle can modify the brain processes to perform a task, then it can also be hypothesized that an intensive mental training or stimulation could provide a greater resilience in facing the neuropathology, possibly recruiting alternate networks. Improving cognitive and functional abilities appears to be a realistic goal in the treatment of Alzheimer’s disease.
According to several authors, it is preferable for rehabilitation to focus on enhancing global cognitive function rather than training distinct domains, since in everyday life, memory, language, executive functions, and reasoning are closely related. Focusing on global functioning would also allow extending the effects of the training to everyday life. It is important to note, however, that since there is still a lack of studies assessing the possibility in AD patients of transferring a specific ability from one context to another, more research in this direction is required before drawing a definite conclusion. Another aspect that needs to be considered is that while it is usually feasible to assess patients’ improvements in a specific domain by using adequate neuropsychological instruments, evaluating global functioning in a standardized way is much more complex. Therefore, both approaches could be usefully attempted, as long as the main objective remains the improvement of the patient’s everyday life.
A very positive aspect of Alzheimer rehabilitation is the actual focus on activities of daily living (ADL), which aims at increasing the length of time patients are able to remain relatively independent, hence limiting caregivers’ burden. It is therefore extremely important to design rehabilitation programs that consider everyday problems that are revealed by patients and caregivers, in order to improve the patients’ quality of life. A detailed planning of the intervention is fundamental for gaining results. Ptak et al. (2010) highlighted the importance of defining target problems and of focusing rehabilitation on concrete goals that should be specific for each subject. This is in agreement with the cognitive goal-oriented rehabilitation approach, which is based on the definition of specific techniques and strategies that are personalized on the basis of the patient’s specific goals and abilities, encouraging the support of caregivers. The involvement of caregivers should be taken into account, not only in the definition of the intervention, but also for supporting goal achievement and in order to have a feedback of the effectiveness of the intervention. The encouraging findings of Clare et al. (2010) regarding the use of cognitive goal-oriented rehabilitation with AD patient should lead to further investigations in this direction.
Being aware of the impact of a lifelong mental stimulation on neuroplasticity and on coping with neuropathology in older age should lead to a greater investment on cognitive interventions for the rehabilitation of the elderly. So far, the question whether there is a difference in the efficacy of cognitive rehabilitation in patients with different years of schooling, different lifelong occupations, and different leisure activities has not been assessed directly. Only in one study (Olazarán et al. 2004), which focused on the benefits of a cognitive-motor stimulation program on patients with mild cognitive impairment (MCI) and mild-to-moderate AD, cognitive response appeared to be higher in individuals with lower education. The authors interpreted this counterintuitive result within the CR framework, according to which, given a similar level of clinical severity, AD pathology would be more advanced in patients with higher education, rendering them in an inferior learning potential condition. Conducting more studies in this direction would be of great importance, allowing the planning of different rehabilitation strategies by taking into account the patients’ different life experiences.
What we can learn from research on CR is the relevance of those particular contexts that facilitate mental engagement, personal growth, and socialization, in promoting more flexible cognitive strategies and therefore more resistance to neuropathology. For this reason, it is advisable to perform rehabilitation within naturalistic settings, where patients have the possibility to spontaneously engage in different activities and socialize with other individuals. A very good example of such context can be found in Adam et al. (2010), describing the management of a patient in a day care center that reproduced a real-life environment, not only aiming at optimizing autonomy and ADL, but also promoting socialization and leisure-oriented activities.
An important aspect that should not be disregarded in the rehabilitation process is the patient’s emotional response. For instance, Kazui et al. (2003) underlined that AD patients can show a memory advantage for emotional stimuli just like healthy subjects. Musical training can certainly be a source of intense emotional experiences, which can enhance motivation. Vanstone and Cuddy (2009) suggested that music could have powerful therapeutic effects if its emotional properties were considered in planning interventions, for instance developing musical cues that could enhance memory by inducing emotions, consequently increasing familiarity with specific aspects of the patient’s daily routine. The positive effects of music on memory are evident in a recent study by Simmons–Stern et al. (2010), showing that sung lyrics were remembered by AD patients better than spoken lyrics.
So far, most rehabilitation interventions have been conceived for patients with mild or moderate AD. A future challenge would be to focus on further stages of the disease. Since in the latest stages of AD communication can be extremely difficult (Miller 1989; Au et al. 1988) interaction should be based on channels other than speech (e.g., gestures, emotional expression). A first attempt in this direction has been recently made with “passive” or “affective” brain–computer interfaces (Nijboer et al. 2009), which aim at extracting involuntary brain signals and utilizing them for human–computer interaction. Adapting affective brain–computer interfaces to Alzheimer patients would allow the restoration of some form of communication even in the latest stages of the disease, leading in the future to new paths for cognitive rehabilitation.