Introduction
Cognitive impairment is the most consistent neurological complication of acquired and degenerative brain disorders. The majority of patients with epilepsy, multiple sclerosis, Parkinson's disease (PD), stroke and head injury have documented cognitive impairments, often early in their disease course Reference Heilman and Valenstein(1). To some extent, this is reflected in self-reported losses (also called subjective memory complaints or meta-memory) but detailed neuropsychological testing is required in order to fully map the degree of impairment (Reference Matotek, Saling, Gates and Sedal2–Reference Gass and Apple5). Yet the degree to which the burden of disease is influenced by changes in cognitive dysfunction is poorly understood. There may be several reasons for this. Cognitive deficits are more appropriately conceived as a cumulative loss of neuronal function supported by specific systems in the brain rather than a single homogenous loss and the impact of heterogeneous losses is difficult to study Reference Mitchell(6). Pre-existing cognitive reserve will influence to what extent new insults impact upon the individual. In addition, cognitive deficits rarely occur in isolation. Separating the effects of cognition from the effects of lack of insight, behavioural problems and other neurological impairments require thorough study design. A further factor is that cognitive deficits are rarely static. They are influenced by natural history of disease, treatment variables and even neuropsychological testing itself.
Many different brain insults are capable of causing a spectrum of deficits from subclinical neuropsychological change through mild cognitive impairment to severe dementia. Compared with classical neurological symptoms such as weakness and sensory loss and established psychiatric complications such as depression or anxiety, the impact of cognitive deficits on the patient and their primary carers has been relatively under investigated. There is accumulating evidence to suggest that patients with neurological disorders are particularly concerned about the preservation of cognitive function, perhaps because of the impact on quality of life (QoL) Reference Fisher, Vickrey and Gibson(7).
The concept of health-related quality of life (HRQoL) is one way to measure the burden of disease. For several decades, clinicians have relied on measurement of impairment (loss), disability (function), handicap (participation) to understand the impact of disease Reference Halbertsma, Heerkens, Hirs, De Kleijn-de Vrankrijker, Dorine Van Ravensberg and Napel(8). However, considerable individual differences in perceived wellbeing, satisfaction and distress are not measured adequately by these constructs Reference Geyh, Kurt and Brockow(9,Reference Gottlieb, Golander, Bar-Tal and Gottlieb10). HRQoL is a much broader concept than impairment or disability but can be difficult to define and measure. The World Health Organisation has revised its taxonomy of International Classification of Impairments, Disabilities and Handicaps (ICIDH) to incorporate QoL (11,12). Higher ratings of QoL are not just important phenomenologically but are also recognised as an important measure of satisfaction with care and adherence with treatment Reference Grosset and Grosset(13). Exploration of the predictors of HRQoL is helped by a clinically applicable definition. Here we use the definition of HRQoL suggested by Mitchell et al. Reference Mitchell, Benito-Leon and Gonzalez(14). That is, the sum of all sources of satisfaction and wellbeing minus all health-related threats (including anticipated threats).
Studies in other areas of medicine reveal multiple predictors of QoL (Box 1). In neurological disorders, there are considerable threats to the self as a result of symptoms in fundamental areas of higher function such as speech and motivation in addition to the threats posed by cognitive decline. Meta-analyses suggest that HRQoL is weakly associated with impairment, moderately correlated with disability and highly correlated with handicap Reference Dijkers(15). Thus, the link between cognitive impairment and burden of disease is unlikely to be simple. HRQoL may be reduced by current cognitive losses as well as the threat of possible future losses. Cognitive domains such as memory, attention, communication, planning, insight and even personal identity could effect satisfaction and daily living in specific ways. Here, we examine possible links between cognitive impairments across neurological disease and impact on HRQoL. Where possible we highlight where specific cognitive domains have been studied in relation to specific areas of HRQoL.
Box 1. Predictors of Poor QoL
Strong predictors
Cognitive impairment
Depression/demoralisation
Lack of autonomy
Lack of support
Chronic Pain
Lack of participation (handicap)
Moderate
Fatigue
Anxiety
Communication difficulties
Bladder and sexual problems
Rapidly progressive disease
Low self-esteem and self-deprecation
Comorbid conditions
Functional loss (disability)
Weak
Long duration of disease
Subtypes of disease
Forced unemployment
Biological disease burden
Older age
Female
Neurological symptoms and signs (impairment)
After Mitchell et al. Reference Mitchell, Benito-Leon and Gonzalez(14).
Methods
Our aim was to systematically review the relationship between cognitive impairment and HRQoL in neurological conditions. We excluded Alzheimer's disease and other dementias that are not preceded by a neurological disorder because the link with HRQoL is likely to be distinct in these conditions and also care is usually provided in a different setting. References for this review were identified by searches of the following dates from inception to October 2009: Web of Knowledge, Scopus, Pubmed, Wiley-Blackwell full text, Sciencedirect full text, Lippincott Williams & Wilkins full text and Ovid full text. The terms ‘quality of life’, ‘disability’, ‘handicap’, ‘participation’, ‘perception’, ‘self-efficacy’, ‘carer or care giver’, ‘health-related quality of life’ and ‘mild cognitive impairment' or ‘cogniti$’ or ‘neuropsycholog* or memory or dysexecutive’ were entered. Articles were also identified through hand searches.
Our search initially identified 4793 articles by abstract search and 1872 by limiting to title search. Of the 1872 articles, there were 421 meeting abstracts, 221 book reviews, 64 editorials, 43 letters and brief communications, and 48 review articles. Of the remaining 1075 primary articles, only 180 examined patients with neurological disease. These formed the basis of this review together with 69 other papers found by citation or hand searches.
QoL in neurological conditions causing cognitive impairment
Brain tumours
Cognitive abnormalities in those with central nervous system (CNS) tumours are common, usually as a direct result of the tumour invasion, but increasingly recognised to occur as a complication of treatment Reference Taphoorn and Klein(16,Reference Brown, Buckner and O'Fallon17). Recent studies, suggest that although both chemotherapy and radiotherapy can cause cognitive deficits, focal radiotherapy in patients with glioma is rarely the main reason for cognitive impairment (Reference Taphoorn and Klein16–Reference Falleti, Sanfilippo and Maruff18). Neuropsychological effects may occur beyond the local tumour site Reference Bartolomei, Bosma and Mein(19). Compared with healthy controls, patients with gliomas have significant reductions in information-processing speed, psychomotor function, attentional functioning, verbal and working memory, executive functioning and HRQoL Reference Klein, Engelberts, Van Der Ploeg and Trenite(20). Cognitive function is also recognised as an independent prognostic factor in the survival of glioma patients Reference Meyers, Hess, Yung and Levin(21,Reference Armstrong, Goldstein, Shera, Ledakis and Tallent22). Not infrequently cognitive deficits are subclinical in those suffering CNS tumours, but tests for attention, memory and word fluency appear to be the most sensitive in detecting subtle dysfunctions. Of particular note, cognitive decline frequently occurs before magnetic resonance imaging (MRI) evidence of disease progression if assessed sensitively Reference Meyers and Hess(23). Unfortunately, there has been little progress in practical methods to test cognitive abnormalities in this population Reference Olson, Chhanabhai and McKenzie(24).
There has been growing interest in the link between cognitive impairment and HRQoL in brain tumours. Subtle cognitive deficits can prevent brain tumour long-term survivors from returning to premorbid autonomy and occupations Reference Giovagnoli and Boiardi(25). In general, when operable, successful resection of the tumour leads to improvement in HRQoL Reference Brown, Maurer and Rummans(26). For those with ongoing complications, the burden of symptoms correlate with HRQoL Reference Osoba, Brada, Prados and Yung(27). Neuropsychiatric symptoms and communication difficulties have been strongly associated with high perceived disease burden Reference Pelletier, Verhoef, Khatri and Hagen(28). Giovagnoli and colleagues compared 94 glioma patients with 24 patients with other chronic neurological diseases and 48 healthy subjects Reference Giovagnoli, Silvani and Colombo(29). The Functional Living Index-Cancer (FLIC) provided HRQoL self-evaluations. In glioma patients, higher FLIC total scores were related to better cognition, physical performances and mood. Compared with healthy subjects, all glioma patients were cognitively impaired and more anxious, and two groups with short duration of recurrence were also more depressed. Glioma and chronic disease patients showed similar FLIC scores and autonomy. Li et al Reference Li, Bentzen and Li(30). examined the relationship between cognition and QoL in 208 patients with brain metastases post-radiotherapy. Cognition was statistically significant correlated with the activities of daily living (ADL) and QoL measured by the Functional Assessment of Cancer Therapy (FACT). Interestingly, scores on all eight neuropsychological subtests deteriorated weeks or months before functional and QoL decline.
Several options are available to measure HRQoL in those with brain tumours from single item tests to complex semistructured questionnaires Reference Mauer, Bottomley and Taphoorn(31). Regarding therapeutic options, there have been a handful of trials of recording HRQoL and cognition Reference Gehring, Sitskoorn and Aaronson(32). One group recently documented improvements in both cognition and HRQoL following an open label trial with an acetylcholinesterase inhibitor Reference Shaw, Rosdhal, D'Agostino, Lovato, Naughton, Robbins and Rapp(33). A second group trialled combined cognitive-rehabilitation and reported high satisfaction with a problem-solving-therapy intervention for 19 patient/caregiver pairs Reference Locke Dona, Cerhan and Wu(34).
Cerebrovascular accidents (stroke)
Stroke is well known to adversely influence current cognitive performance Reference Vermeer, Prins, Den Heijer, Hofman, Koudstaal and Breteler(35), HRQoL Reference Paul, Sturm, Dewey, Donnan, Macdonell and Thrift(36) and ultimately mortality Reference Melkas, Oksala and Jokinen(37). Stroke survivors with frequent complications report poorer HRQoL in the area of memory and thinking, and those with an ischaemic stroke and concordance reported poorer communication Reference Nichols-Larsen, Clark, Zeringue, Greenspan and Blanton(38). Two large studies have recently reported rates of post-stroke cognitive impairment of approximately 40–55% Reference Madureira, Guerreiro and Ferro(39,Reference Patel, Coshall and Rudd40). The relative risk of dementia after stroke is five-fold compared with that in aged-matched controls Reference Hénon, Durieu and Guerouaou(41,Reference Tatemichi, Paik and Bagiella42).
What then is the link with HRQoL? Certainly extent of the lesion and depression are associated with poor post-stroke HRQoL but other factors are also important Reference Fatoye, Komolafe and Eegunranti(43,Reference Moon, Kim, Kim, Won and Kim44). Kwa et al. Reference Kwa, Limburg and De Haan(45) conducted one of first studies to specifically examine cognitive aspects in 129 patients with ischaemic stroke. Cognitive function was assessed with the Cambridge cognitive battery (CAMCOG) and HRQoL with a visual analogue scale (VAS). Regression analysis revealed that disturbed global functional health, larger volume of infarcts and severity of aphasia were significantly independent explanatory factors for poorer HRQoL. Thus, in a relatively high functioning group, there was no significant impact of cognitive impairment on the patients' HRQoL, beyond the effect on other domains. Jonkman et al. Reference Jonkman, De Weerd and Vrijens(46) found similar results first observing that QoL improved to a modest extent in the period 3–12 months after the stroke and that change in QoL was correlated with depression, neurological deficit, but not with cognitive disturbances. However, Hochstenbach et al. Reference Hochstenbach, Anderson, Van Limbeek and Mulder(47) used more comprehensive neuropsychological testing in 164 stroke patients and found HRQoL was associated with deficits on specific tests measuring spatiotemporal and/or sequential aspects of behaviour. Poor HRQoL was more likely if patients had a poor result on the Trail making Test B. In a follow-up study 2 years later, significant improvements across time were noted for all cognitive domains and interestingly the largest improvement was found in the attentional domain; the least, in the memory domain Reference Hochstenbach and Mulder(48). A small subset of patients accounted for the significant cognitive improvement because most patients showed no improvement or declined. In a small group of 35 stroke patients after a stroke, Kauhanen et al. Reference Kauhanen, Korpelainen and Hiltunen(49) found that infarct volume, aphasia, impaired motor function and impaired cognitive function were linked with poorer QoL, but again depression was the most significant. Halari et al. Reference Hilari, Wiggins, Roy, Byng and Smith(50) looked at 95 participants with chronic aphasia (>1 year) from three different sites. Emotional distress, involvement in activities, extent of communication disability and number of comorbid conditions explained 52% of the variance in HRQoL. Nys et al. Reference Nys, Van Zandvoort and Van Der Worp(51) examined 143 patients within the first 3 weeks post-stroke using the Montgomery Asberg Depression Rating Scale and Stroke-Specific Quality of Life Scale. Cognitive impairment (especially problems with visual perception and construction), increasing age and functional dependence predicted a reduced QoL.
Unlike many neurological conditions, here have been several serious attempts to validate instruments to measure stroke-related cognitive change but as yet no consensus about which test is optimal for clinical practice (Reference Randhawa, Walterfang and Miller52–Reference Adunsky, Fleissig and Levenkrohn55). Regarding interventions some early studies have reported positive effects of cognition and HRQoL with various rehabilitation packages Reference Olsson, Sunnerhagen and Stibrant(56).
Epilepsy
Cognitive deficits in patients with epilepsy can be static (for instance, when deficits are associated with hippocampal sclerosis), episodic (for example, when deficits are post-ictal) or progressive (memory deficits in recurrent status epilepticus) Reference Lutz and Helmstaedter(57). Cognitive deficits associated with the adverse effects of medication are potentially reversible, whereas deficits caused by structural changes (including epilepsy surgery) are not.
Although there are numerous studies of QoL in epilepsy, the potential influence of cognitive impairment on QoL has received relatively little attention and with no allowance for epilepsy type or treatment received. Patients with partial epilepsy (even those able to maintain regular employment) are now recognised to be at risk for cognition impairment Reference Andersson-Roswall, Engman and Samuelsson(58). Previously there was also an assumption that poor HRQoL must be attributed directly to seizure frequency or medication side-effects. In a landmark study, Perrine et al. Reference Perrine, Hermann and Meador(59) looked at 257 patients with various types of epilepsy from 25 epilepsy centres in the United States. Mood, psychomotor speed, verbal memory and language correlated significantly with selected scales of the QoL in Epilepsy-89 inventory and were predictive of overall QoL. Mood was the strongest predictor in regression analysis. In a study of 65 patients with temporal lobe epilepsy (TLE), regression analysis showed that overall QOLIE-89 score was predicted by the factor Mood and Questionnaire of Memory Efficiency score. The QOLIE-89 factor cognition was predicted by the Questionnaire of Memory Efficiency score and the memory, mental speed, perception and praxis factors of the neuropsychological test battery Reference Giovagnoli and Avanzini(60). Engelberts et al. Reference Engelberts, Klein and Van Der Ploeg(61) looked at 56 outpatients with partial epilepsy who were compared with 56 matched healthy controls. Mental aspects of HRQoL such as fatigue were lower, whereas physical functioning was unaffected. These patients also expressed reductions in mental functioning as indicated by low self-perceived cognitive functioning. Duration of epilepsy, seizure type, seizure frequency, localisation, years on carbamazepine and carbamazepine dosage were not related to cognitive functioning or HRQoL. In a further study by Engelberts et al. Reference Engelberts, Klein and Van Der Ploeg(62) looked at cognition (attention, memory and processing speed) in relatively stable outpatients with partial epilepsy treated with carbamazepine. A lower speed of information processing affecting everyday life functioning was found. Lower levels of self-perceived neuropsychological functioning were reported, whereas HRQoL was unimpaired relative to healthy controls. Piazzini et al. Reference Piazzini, Beghi and Turner(63) examined a large group of 815 patients recruited from 24 secondary and tertiary Italian centres for the care of epilepsy using the Epi-QoL. The main predictors included gender, seizure frequency, prognostic categories, number of medications, comorbidity, presence of cognitive impairment, psychiatric disturbances and disability. Meneses et al. Reference Meneses, Pais-Ribeiro and Silva(64) studied only 71 patients with focal epilepsy of moderate severity but with comprehensive neuropsychological testing. Attentive matrices and semantic fluency predicted several aspects of HRQoL with the exception of social functioning.
Several groups have examined the effects of epilepsy surgery on cognitive performance and others on HRQoL but not in the same study. Those who achieve an improvement in seizure control usually report improved HRQoL Reference Selai, Elstner and Trimble(65,Reference Lowe, David and Kilpatrick66). The effect is particularly marked where an individual regains independence and social function Reference Aydemir, Ozkara, Canbeyli and Tekcan(67). What then is the effect on cognition? Resective procedures, especially in the dominant hemisphere, usually cause memory impairment Reference Ojemann and Dodrill(68,Reference Joo, Han and Lee69). Both anterior temporal lobectomy and amygdalohippocampectomy have been shown to worsen verbal learning and memory Reference Helmstaedter, Reuber and Elger(70). Preoperative ability status, age at the time of surgery and the extent of the resection of functioning tissues are the most important factors determining post-operative memory reserve Reference Davies, Bell, Bush and Wyler(71,Reference Helmstaedter and Elger72). However, this detrimental effect has to be balanced against positive cognitive benefits of seizure control. In patients with improved seizure control, there was evidence of recovery of memory and non-memory functions Reference Helmstaedter, Kurthen, Lux, Reuber and Elger(73).
Head injury (traumatic brain injury)
Cognitive deficits are one of the most important complications of head injury, being closely related to global severity Reference Wilson, Pettigrew and Teasdale(74,Reference McCarthy, Dikmen, Langlois, Selassie, Gu and Horner75). Some form of memory disturbance almost always accompanies moderate or severe head injuries, and not infrequently mild head injury as well Reference Belanger, Curtiss, Demery, Lebowitz and Vanderploeg(76). Yet, severe cognitive deficits occur in less than 5% of head injuries Reference Kemp, Goulding and Spencer(77). Reduced attention, memory deficits (anterograde amnesia) and word-finding difficulties are also seen but are typically transient. Risk factors for cognitive deficits are penetrating injuries, head injuries in elderly patients and lesions to frontal or dominant temporal lobes Reference Frencham, Fox and Maybery(78). Recovery seems to occur more quickly in domains including memory, verbal abstraction and manual dexterity than in other domains Reference Christensen, Colella and Illness(79).
Cognitive deficits after head injury have an important influence on long-term outcome Reference MacKenzie, McCarthy and Ditunno(80). Return to work has been studied extensively in this regard. Early cognitive deficits affecting attention, memory and orientation have been used as predictors of functional recovery and return to work Reference Dawson, Levine, Schwartz and Stuss(81,Reference Atchison, Sander and Struchen82). However, relatively few studies have looked specifically at QoL as an outcome Reference Dijkers(83). Lannoo et al. Reference Lannoo, Van Rietvelde, Colardyn, Lemmerling, Vandekerckhove, Jannes and De Soete(84) studied 158 patients with severe head injuries in comparison with group of 32 patients with traumatic injuries to parts of the body. Head injury patients had poorer scores on the sickness impact profile (SIP) scales for mobility, intellectual behaviour, communication, home management, eating and work. Tate and Broe Reference Tate and Broe(85) examined the long-term predictors of disability in 70 individuals with varying levels of disability after severe traumatic brain injury. Examination of specific domains of psychosocial functioning (occupational activities, interpersonal relationships and independent living skills) revealed different patterns of significant predictor variables. Occupational activity was best predicted by physical impairment and memory, whereas interpersonal relationships were predicted by time post-trauma, cognitive processing speed and behaviour. Independent living was predicted by physical impairment and behavioural regulation and memory deficits. Wood and Rutterford Reference Wood and Rutterford(86) looked at multidimensional predictors of long-term outcome in 131 participants more than 10 years after injury. In this study, impairment in working memory did not predict QoL.
Occupational status is one objective measure of global functioning that correlates with QoL ratings in a variety of medical disorders (Reference Blalock, McDaniel and Farber87–Reference Orbon, Schermer and Van Der Gulden89), including head injury Reference Johansson and Bernspang(90). A number of independent groups have examined cognitive performance as a predictor of return to work (Reference Ownsworth and McKenna91–Reference Green, Colella and Hebert93). In general, the factors most consistently associated with employment outcome included pre-injury occupational status, functional status at discharge, global cognitive functioning, involvement in rehabilitation services and emotional/behavioural complications. Specific cognitive domains may have particular importance (Reference Green, Colella and Hebert93–Reference Simpson and Schmitter-Edgecombe95). Even in the case of mild traumatic brain injury, verbal memory, verbal fluency and a speed test of planning and strategy are predictive of work status 3–15 months later Reference Drake, Gray, Yoder, Pramuka and Llewellyn(96). In brain injury, there are well-designed randomised trials of neuropsychological rehabilitation showing that treatment can influence long-term outcomes including HRQoL (Reference Ponsford, Olver, Ponsford and Nelms97–Reference Cicerone, Mott and Azulay100).
Huntington's disease
Changes in executive function and memory function are often noticeable at least 2 years before the clinical development of the condition in at-risk carriers Reference Paulsen, Zhao and Stout(101). Most, but not all, studies have found that the severity of cognitive impairment is determined by the number of CAG repeats. Deficits are progressive, and the point prevalence of dementia in definite Huntington's disease is somewhere between 30 and 50%, depending on the sample and instrument used.
QoL has been poorly studied in this condition, particularly in relation to cognitive change. Helder et al. Reference Helder, Kaptein and Van Kempen(102) assessed 77 patients with a clinically confirmed diagnosis of Huntington's disease (HD) using the SIP. Total motor score, MMSE scores and the duration of HD were significantly correlated with patients' scores on the SIP, and predicted a significant amount of variance of the physical dimension of the SIP, but not of the psychosocial dimension. Ho et al. Reference Ho, Gilbert and Mason(103) studied 70 patients with HD using the Unified Huntington's Disease Rating Scale (UHDRS) motor, Beck Depression inventory and SF36. Poor HRQoL was associated with higher levels of depressive mood and lower functional capacity but interestingly motor symptoms and cognitive function were not particularly influential.
Regarding assessment, several small-scale studies have proposed certain generic instruments to detect cognitive change in HD Reference Salmon, Kwoonyuen and Heindel(104,Reference Paulsen, Stout and TawfikReedy105). Relatively little has been published concerning the value of different HRQoL instruments in Huntington's disease Reference Ho, Robbins, Walters, Kaptoge, Sahakian and Barker(106,Reference Higgins, Nickerson, Molho and Factor107) and there have been no therapeutic trials in this area.
Motor neuron disease
Significant neuropsychological impairment is present in at least one-third of all sufferers of classical MND Reference Neary, Snowden and Mann(108). Until recently, MND was seen as a pure disorder of the motor system with a low incidence of cognitive impairment. Increasingly, it is recognised that patients with MND without dementia often manifest fronto-temporal type cognitive impairment. Indeed MND and fronto-temporal dementia can overlap in both familial and sporadic cases of MND Reference Abrahams, Leigh and Goldstein(109). Selective cognitive impairment in the form of verbal fluency deficits, most likely indicating executive dysfunction, appears relatively early on in the course of the disease, and language functions may become vulnerable as the disease progresses. Dementia occurs in 5–10% of sufferers, with an increased risk of frontal lobe dementia in the late stages of disease Reference Barson, Kinsella, Ong and Mathers(110). Authors have looked for biological markers of cognitive decline in MND. Provisional work suggests that on PET functional imaging greater impairment of medial temporal lobe activity occurs in MND with fronto-temporal dementia compared with those without dementia Reference Garraux, Salmon and Degueldre(111). Nevertheless, cognitive impairment is not an invariable feature of MND and can be difficult to detect Reference Duque, Parama and Borges(112).
The relationship between cognitive decline in MND and perceived QoL has rarely been studied. Of note, in some conditions, cognitive impairment appears to have a modest, but significant impact on patients' desire for hastened death Reference Pessin, Rosenfeld, Burton and Breitbart(113). In a study of 31 patients with MND, QoL scores were found to correlate positively with the existence of confiding and emotional support and also correlated negatively with the presence of self-rated everyday cognitive difficulties but not with affective state Reference Goldstein, Atkins and Leigh(114). As with other neurological disorders, both fatigue and depression are associated with poorer QoL in subjects with MND Reference Lou, Reeves, Benice and Sexton(115,Reference Kubler, Winter, Ludolph, Hautzinger and Birbaumer116). Instruments to measure QoL are available but their validation is unclear Reference Epton, Harris and Jenkinson(117).
Multiple sclerosis
When examined carefully, specific cognitive deficits can be seen in over 50% of patients, often in the earliest phases of disease Reference Rogers and Panegyres(118,Reference Chiaravalloti and DeLuca119). However, up to half of such cases are overlooked when using a routine examination alone Reference Brassington and Marh(120), and personal accounts may be also be revealing Reference Shevil and Finlayson(121). Even in patients with short disease duration of less than 2 years, discrete impairment of cognitive function may be found in up to 60% on neuropsychological testing without impacting activities of daily living Reference Lyon-Caen, Jouvent and Hauser(122,Reference Ruggieri, Palermo, Vitello, Gennuso, Settipani and Piccoli123).
Most (Reference Shawaryn, Schiaffino, LaRocca and Johnston124–Reference Gold, Schulz, Monch, Schulz and Heesen127) but not all Reference O'Connor, Lee, Ng, Narayana and Wolinsky(128) studies document an association between cognitive deficits and lowered HRQoL. Cutajar et al. Reference Cutajar, Ferriani and Scandellari(126) evaluated patients with MS in a cross-sectional study. A relationship was found between both memory impairment and executive function and HRQoL. Gold et al. Reference Gold, Schulz, Monch, Schulz and Heesen(127) compared 80 MS patients with cognitive dysfunction with 107 unimpaired patients, separated on the basis of the symbol digit modalities test (SDMT). Cognitively impaired patients showed significantly higher depression and anxiety as well as lower HRQoL. Shawaryn et al. Reference Shawaryn, Schiaffino, LaRocca and Johnston(124) used the paced auditory serial addition test (PASAT) as a measure of cognitive impairment in 90 patients with MS. Information-processing speed predicted mental and emotional aspects of HRQoL on the Medical Outcomes Trust's Short Form-36 (SF-36) and both the EDSS and the PASAT predicted depression. Benedict et al. Reference Benedict, Wahlig and Bakshic(129) found that cognitive decline predicted mental components of HRQoL (in unadjusted analysis) and ability to maintain employment. Most recently, Benito-Leon et al. Reference Mitchell, Benito-Leon and Rivera-Navarro(130) examined all degrees of cognitive impairment using neuropsychological testing as well as the Clock Drawing Test and MMSE screening instruments in 191 MS patients. After controlling for depression comprehensive (but not simple) ratings of cognition significantly contributed to poor HRQoL. In addition, all degrees of cognitive impairment, severity of fatigue and higher physical disability were independent predictors of low FAMS instrument total scores. Our search only identified one trial linking rehabilitation with improvements in HRQoL Reference Patti, Ciancio and Reggio(131). Patti et al. Reference Patti, Ciancio and Reggio(131) found improvements in HRQoL in those randomised to an intensive 6-week rehabilitation package.
Parkinson's disease
Cognitive impairment is increasingly recognised as an important feature of PD Reference Martinez-Martin(132). A substantial proportion develop Lewy body dementia Reference Mindham and Hughes(133,Reference Aarsland, Ballard and Glenda Halliday134). Specific cognitive domains are affected even in early disease stages Reference Bruck, Kurki, Kaasinen, Vahlberg and Rinne(135) and clinical criteria have been suggested Reference Emre, Aarsland and Brown(136). Common problem areas include executive performance, which is more severely affected than visuospatial function, which in turn is more severely affected than memory and language.
Schrag et al. Reference Schrag, Jahanshahi and Quinn(137) examined the factors that influence HRQoL using the specific Parkinson's disease Questionnaire (PDQ39). The factor most closely associated with disruptions in HRQoL was the presence of depression (explaining 54% of the variance). Disability, postural instability, a history of falls or of gait difficulties and cognitive impairment (as determined by MMSE) also contributed to poor HRQoL. In a small study, Croyle et al. Reference Croyle, Fields, Lyons, Straits-Troster, Koller, Pahwa and Troster(138) showed that neuropsychological deficits were linked with pool QoL. Also in a community study of 114 patients with idiopathic PD, Weintraub et al. Reference Weintraub, Moberg, Duda, Katz and Stern(139) showed that increasing severity of depression and worsening cognition on the MMSE were associated with greater disability using the UPDRS ADL score, accounting for 37% of the variance in disability. Unfortunately, QoL per se was not examined. Klepac et al. Reference Klepac, Trkulja and Relja(140) studied 124 consecutive PD patients without dementia based on DSM-IV criteria and Mini Mental State Examination. In multivariate analysis, after multiple adjustments better cognitive performance was independently associated with better HRQoL.
Treatments for cognitive decline are showing some promise in PD and may influence QoL Reference Emre, Aarsland and Albanese(141). A rehabilitation program of 6 weeks, including both motor and cognitive training, was applied to 20 patients affected by PD in the early stages, presenting with mild cognitive deficits, but no dementia. At the end of the scheduled sessions, the patients showed a significant improvement in verbal fluency, logic memory and Raven's matrices tests, when compared with baseline Reference Sinforiani, Banchieri, Zucchella, Pacchetti and Sandrini(142).
Conclusions
A spectrum of cognitive impairments complicates the course of a substantial proportion of patients with all the neurological conditions discussed in this systematic review. In the case of epilepsy, Huntington's disease, motor neuron disease, multiple sclerosis and PD deficits accumulate early in the disease course, sometimes pre-clinically. Although much attention has focussed on dementia, there is increasing recognition of the importance of a broad spectrum of impairments, such as those rated as subclinical, minimal or mild Reference Albert, Michaels and Padilla(143).
Given that cognitive difficulties are a key part of neurodegenerative and acquired brain disorders, it is surprising that they often undetected and untreated Reference Chodosh, Petitti, Elliott, Hays, Crooks, Reuben, Buckwalter and Wenger(144). Even frank dementias often go unrecognised for many years Reference Valcour, Masaki, Curb and Blanchette(145). Yet cognitive symptoms predict future mortality in healthy and dementia populations Reference Feil, Marmon and Unutzer(146,Reference Cacciatore, Abete, De Santis, Longobardi, Ferrara and Rengo147) and have prognostic value in other medical disorders Reference Henon, Durieu, Lebert, Pasquier and Leys(148). Moreover, cognitive deficits are an important outcome variable in themselves. This systematic review shows that in many neurological conditions mild, moderate and severe cognitive change is perceived as a significant burden by both patients and carers. In other areas of medicine, perceived memory difficulties, poor concentration and impaired problem solving are important predictors of low HRQoL Reference Kiessling and Henriksson(149), as well as the ability to continue successful employment Reference Kiessling(150,Reference Rabkin, McElhiney, Ferrando, Van Gorp and Hsing151). However, losses in specific domains appear to affect global outcomes in different ways. For example, Artero et al. Reference Artero, Touchon and Ritchie(152) examined predictors of decline in disability over a 3-year period in patients with mild cognitive impairment. Decline in visuospatial performance on cognitive tests was the most significant predictor of decline in ability to dress, bathe, use the telephone or retain mobility. Decline in attention also reduced the capability of independent toileting, and a decline in language was associated with poor telephone use. Memory impairment alone did not have a large effect. Similarly, after head injury return to work may be best predicted by memory problems, whereas interpersonal relationships are more closely tied with cognitive processing speed and behaviour Reference Possl, Jurgensmeyer, Karlbauer, Wenz and Goldenberg(153). Cognitive difficulties can also adversely influence personal and social interactions. For example, the inability to predict the actions of others and remember previous interactions is likely to erode personal relationships and thereby threaten social support. In several areas (particularly brain injury), neuropsychological rehabilitation programmes have helped patient and caregiver satisfaction and improved HRQoL.
This systematic review is limited by the paucity of research in many neurological conditions. Relatively few studies have examined the progression of deficits or specific cognitive domains in relation to HRQoL. There is also considerable uncertainty about what degree of severity of cognitive change is necessary in order to produce a measurable effect on global wellbeing. Further research is required to examine vulnerability and reliance factors in HRQoL and in particular the role of social support and insight. Nevertheless, we suggest that a thorough cognitive examination should be part of routine clinical practice in those caring for neurological patients.
