- Email: [email protected]
Cerebral correlates of cognitive reserve
Psychiatry Research: Neuroimaging 247 (2016) 65–70
Contents lists available at ScienceDirect
Psychiatry Research: Neuroimaging journal homepage: www.elsevier.com/locate/psychresns
Review article
Cerebral correlates of cognitive reserve Lawrence J Whalley a,b,n, Roger T Staff a,c, Helen C Fox a, Alison D Murray a a b c
Aberdeen Biomedical Imaging Centre, College of Life Sciences and Medicine, University of Aberdeen, Aberdeen AB25 2ZD, UK Institute of Applied Health Sciences, University of Aberdeen, Aberdeen, UK NHS Grampian, Foresterhill, Aberdeen, UK
art ic l e i nf o
a b s t r a c t
Article history: Received 9 October 2015 Accepted 13 October 2015 Available online 19 October 2015
Cognitive reserve is a hypothetical concept introduced to explain discrepancies between severity of clinical dementia syndromes and the extent of dementia pathology. We examined cognitive reserve in a research programme that followed up a non-clinical sample born in 1921 or 1936 and IQ-tested age 11 years in 1932 or 1947. Structural MRI exams were acquired in about 50% of the sample from whom a subsample were recruited into an additional fMRI study. Here, we summarise findings from seven interrelated studies. These support an understanding of cognitive reserve as a balance between positive life course activity-driven experiences and the negative effects of brain pathologies including cerebrovascular disease and total and regional brain volume loss. Hypothesised structural equation models illustrate the relative causal effects of these positive and negative contributions. Cognitive reserve is considered in the context of choice of interventions to prevent dementia and the opposing effects of cerebrovascular disease and Alzheimer like brain appearances. Crown Copyright & 2015 Published by Elsevier Ireland Ltd. All rights reserved.
Keywords: Cognitive reserve Childhood IQ White matter hyperintensities Functional MRI Cortical complexity Structural equation model Cognitive aging Alzheimer's disease
Contents 1.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1. Cerebral correlates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2. Cognitive reserve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. The Aberdeen birth cohort studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. The studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Cognitive reserve and brain volume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Intelligence, cognitive reserve and brain aging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3. An fMRI study of successful cognitive aging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4. Cortical complexity and cognitive aging. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5. Childhood socioeconomic status influences adult hippocampal size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6. Is there a balance between cognitive reserve, cerebrovascular disease and Alzheimer's disease? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.7. Do childhood socioeconomic circumstances influence the occurrence of brain hyperintensities? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Comment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1. Introduction Population
trends
in
age
structure
accelerated
rapidly
n Corresponding author at: Aberdeen Biomedical Imaging Centre, College of Life Sciences and Medicine, University of Aberdeen, Aberdeen AB25 2ZD, UK. E-mail address: [email protected] (L. Whalley).
65 66 66 66 66 66 66 66 67 67 67 69 69 69 70 70
throughout the 20th century. As numbers of old people increased, so concerns grew that diseases associated with aging would increase in prevalence and potentially jeopardise health care provisions for the elderly. Dementia is a major cause of disability and death affecting about 5% of all those over age 65 years and about 20% of those aged 85 years or more. Although recent reports of falling prevalence of dementia give rise to cautious optimism that
http://dx.doi.org/10.1016/j.pscychresns.2015.10.012 0925-4927/Crown Copyright & 2015 Published by Elsevier Ireland Ltd. All rights reserved.
66
L.J Whalley et al. / Psychiatry Research: Neuroimaging 247 (2016) 65–70
dementia may prove somewhat less of a burden on health services, effective dementia prevention strategies remain a major priority in the 21st century. Disappointed by the results of recent trials of anti-amyloid therapies in Alzheimer's disease (AD), alternative approaches to dementia prevention are now widely discussed. Attention has turned to possible explanations of decreased AD prevalence detected in surveys in Sweden (Qiu et al., 2013) and England (Matthews et al., 2013) and comparable data from Holland and the US. From these reports, two broad themes emerged. The first concerns comparisons between recent reduced dementia prevalence and established downward trends in heart-disease and stroke (McGovern et al., 1996; Soros et al., 2012). The second, draws on reports of the benefits for cognitive aging and possible dementia protection that can be linked to lifestyles that are cognitively demanding, socially engaged and physically active (Angevaren et al., 2008; Fratiglioni and Qiu, 2011; Barnes et al., 2013). In this review, we describe our approach to the concept of cognitive reserve. Our aim is to make tractable the investigation of cognitive reserve with the intention of identifying those components that could enhance its protective or compensatory potential and inform the design of studies to prevent or delay dementia onset.
2. Methods 2.1. The Aberdeen birth cohort studies Since 1998, we have followed up local community residents who at age about 11 years were at school in Aberdeen and sat a group-administered IQ-type test. Full details are given elsewhere (Whalley et al., 2011). Briefly, the Aberdeen birth cohort studies were started in 1998 when Scottish Mental survey archives of the Scottish Council for Research in Education were re-discovered and permissions obtained to follow-up survivors born in 1921 or 1936 and then aged about 77 or 64 years and who had entered (or were about to enter) the age of greatest risk for Alzheimer's disease (AD). Sources of attrition from the study, exposures to childhood adversity, nutritional, genetic and life style factors of possible relevance to extent of age-related cognitive decline and the timing of onset of dementia are given elsewhere (Whalley et al. 2011). By 2014, the feasibility of following up more than 75% of Scottish Mental Survey survivors living in the Aberdeen area without dementia was well-established, dementia ascertainment to age about 88 years was completed in the 1921 birth cohort and was underway in the 1936 born cohort (by then aged about 78 years). Structural MRI exams were introduced in 2001 and, by 2008, about 50% of the total sample of 791 participants has undergone at least one MRI. These databases are available to other bone fide research groups wishing to test specific hypotheses that may either replicate their own findings or make best use of the data collected in the Aberdeen studies (contact [email protected]).
3. The studies 3.1. Cognitive reserve and brain volume
1.1. Cerebral correlates The cerebral cortex is a complex layered and much folded bilateral structure that makes up the cerebral hemispheres. Agerelated changes in the cerebral cortex are multi-factorial and affect neurones and glia, the cerebral vasculature and the white matter tracts. In life, these changes are detectable using advanced brain imaging techniques that accurately measure brain structures, vascular pathologies (including small vessel disease) and regional brain metabolic differences. Our studies on cognitive reserve have largely relied on structural magnetic resonance imaging (MRI) and, to a much lesser extent, functional MRI (fMRI) to reveal differences in blood oxygen level detection (BOLD) responses in comparisons between experimental conditions. 1.2. Cognitive reserve Quantification of cognitive reserve depends upon the precise measurement of cognitive performance. A wide range of cognitive domains makes up overall cognitive function and shares common variance (a general factor ‘g’) that explains the extent to which domains are inter-correlated. In our studies we have selected a small number of standardised cognitive tests that capture verbal learning (auditory verbal learning test, AVLT) and spatial ability (Block Design, BD); non-verbal reasoning, Raven's Standardised Progressive Matrices (RPM); mental speed (Digit Symbol, DS) and verbal fluency (Guilford's Uses of Common Objects). Baseline childhood IQ was taken from historical records of testing of children using the Moray House Test (Deary et al., 2009). Age-related cognitive change was estimated from the difference between observed current cognitive performance in late adulthood and cognitive performance predicted from Moray House Test scores converted to standard IQ-type scores (mean 100, standard deviation 15). It was hypothesised that cognitive reserve would be a major positive influence on individual differences between observed current cognitive performance and expected performance predicted by childhood IQ. Negative influences include age-related brain pathology (brain atrophy and brain lesion load thought to be of cerebrovascular origin).
In our first investigation into the concept of cognitive reserve, we explored the possibility that reserve was influenced by head size (as a proxy for lifetime maximum brain size), duration of education, and occupational attainment in 92 volunteers born in 1921 and whose childhood IQ age 11 years was available. We performed neuropsychological tests of verbal memory, spatial ability, non-verbal abstract reasoning, verbal fluency and mental speed at age about 79 years and within 3 months conducted MRI exams. The fraction of total intracranial volume occupied by brain at age 79 years was used as a measure of brain shrinkage (atrophy) and density of white matter hyperintensities as a measure of cerebrovascular disease (burden). The results showed that education and occupational attainment but not total intracranial volume (a proxy for maximum brain size) contribute in this preliminary study to retention of cognitive performance from age 11 to 79 years (Staff et al., 2004). Because we had baseline childhood IQ scores, we could adjust age 79 cognitive test scores for baseline and use general linear modelling of covariance to test cognitive reserve hypotheses. Separate models were tested for verbal memory and non-verbal reasoning. The simplest model of cerebral reserve is that it is a passive property that protects the individual from age-related cognitive decline. This line of reasoning implies that a person with a larger brain is capable of withstanding a greater degree of age-related pathology and would decline less than a person with a smaller brain. An active model of cognitive reserve can be tested in a similar way after adding education as a likely contributor to verbal memory but not to non-verbal reasoning. These models of active and passive reserve were first suggested by Stern (2002). 3.2. Intelligence, cognitive reserve and brain aging In a later analysis, using the same MRI data, and after accounting for childhood IQ, we explored relationships between specific cognitive domains, the general factor (‘g’) that explained variance shared between test and brain volumes. We hypothesised that brain aging would be correlated with non-uniform reductions in brain volumes and that these would be associated with scores on specific cognitive tests (Staff et al., 2006). As predicted, we found links between specific cognitive tests and brain volumes. An
L.J Whalley et al. / Psychiatry Research: Neuroimaging 247 (2016) 65–70
association was also found between brain volumes and the general factor ‘g’. After removing the influence of ‘g’ from each of the specific cognitive test scores, no remaining significant associations were found between brain volumes and specific cognitive domain test scores. We concluded that cognitive aging across multiple domains of cognitive function is associated with brain volume differences but that – in this small sample of older adults without dementia – we did not find evidence that specific components of cognitive aging were associated with smaller brain volumes. We found preliminary evidence that the white matter tracts of the left (dominant) hemisphere were reduced in size and that this correlated with overall cognitive aging. We tentatively concluded that these reductions were causally linked to cognitive aging but since ours was a cross-sectional and not a longitudinal study, we cannot be sure that these did not pre-date the onset of cognitive aging. An alternative explanation would be that these white matter tracts reflect a pre-existing “passive” component of cognitive reserve that allows an individual to adjust more effectively to brain aging. Other studies have found these white tracts are directly linked in size to chronological age and this would support their causal rather than a reserve role in cognitive aging (Gunning-Dixon and Raz, 2000). 3.3. An fMRI study of successful cognitive aging The Aberdeen 1936 birth cohort was recruited aged about 64 years. At age 70 years, a small sub-sample participated in an fMRI study. At this stage in the development of our research programme, we were interested in neurobiological mechanisms that could underpin (i.e. be “fundamental”) to differences in rates of cognitive aging. We anticipated that at least in part these biological findings would prove relevant to understanding the neurobiology of cognitive reserve (Whalley et al., 2004). Briefly, the design of the experiment was as follows. We compared two groups of non-demented 70-year-olds who, at age 11, had relatively similar general cognitive ability but who, in old age, had diverged, with one group demonstrating relatively successful (“sustainers” N ¼25), and the other one unsuccessful, cognitive ageing (“decliners” N ¼15). We examined whether, in older people with relatively successful cognitive ageing, their BOLD activation– deactivation pattern while they performed an inspection time task was more or less similar to those of healthy younger individuals than older people with relatively unsuccessful cognitive ageing (Waiter et al., 2008). Inspection time is a two-alternative forced choice, backward masking test of the speed of the early stages of visual information processing. Inspection time has a well-established, significant association with higher cognitive abilities. The group of cognitive sustainers showed a pattern of BOLD activation–deactivation in response to inspection time stimulus duration differences that was similar to a healthy young sample (Deary et al., 2004). The group of cognitive decliners lacked these clear neural networks. The relative preservation of complex reasoning skills in old age may be associated with the preservation of the neural networks that underpin information processing in youth. We concluded that retention of youthful neural networks by this small group of “sustainers” who had aged relatively successfully to that point provided a neurobiological basis for their retention of complex reasoning skills. These advantages may be generalisable to other cognitive domains and prove relevant to understanding the neurobiological basis of cognitive reserve. Age-related cognitive performance does not invariably decline and some individuals will improve in late adulthood. One explanation, with strong empirical support, is that the slowing of information processing speed is an important influence on cognitive aging. The neurobiology of age-related changes in processing speed is poorly understood. Sampling successful agers using a
67
strategy based on complex tests of reasoning, showed brain patterns of activity that resemble younger people's during a visual information processing task. 3.4. Cortical complexity and cognitive aging We recognised that in order to understand how neural networks performed in supporting cognition in the face of brain aging, our next step should aim to quantify structural differences in the complexity of neural networks. We hypothesised that successful cognitive agers would have more complex cerebral cortices than individuals who declined. Previously, we had rejected the “more is better” hypothesis of successful cognitive aging so began to explore alternative approaches to cortical structure. McDaniel (2005) had presented a meta-analysis of reports on brain size and general mental ability and concluded that brain size made a modest but significantcontribution to IQ. The cerebral cortex is a fractal like structure made of repeated elements each of which resembles the structure as a whole. Cortical fractal structure can be characterised by a single numerical value – the fractal dimension (FD) – that summarises the irregularity and reproducibility at different scales of (1) the external cortical surface and (2) the internal boundary between sub-cortical grey and white matter. Figs. 1 and 2 illustrate how FD increases with increasing complexity of the surface under investigation. This is easily pictured in comparisons between coastlines (Fig. 1) though less readily seen in cortical surfaces (Fig. 2). In the development and aging of the human brain, there is increasing cortical complexity from early foetal life (Shyu et al., 2010) and into adulthood (Takahashi et al. 2004) until in late life when decreasing cortical complexity is seen (King et al., 2009). We decided to examine associations between the FD of white matter and cognitive changes across the life course in the absence of detectable brain disease (Mustafa et al., 2012). The FD was calculated from segmented cerebral white matter MR images in 217 subjects aged about 68 years. They were sampled from the Aberdeen 1936 birth cohort described above for whom we had archived IQ scores from age 11 years. Cognitive test scores of fluid and crystallised intelligence were obtained at the time of MR imaging. Significant differences were found (in intracranial volume, brain volume, white matter volume and Raven's Progressive Matrices score) between men and women at age 68 years and novel associations were found between FD and measures of cognitive change over the life course from age 11 to 68 years. Those with greater FD were found to have greater than expected fluid abilities at age 68 years than predicted by their childhood intelligence and less cognitive decline from age 11 to 68 years (Table 1). These results are consistent with other reports that FD measures of cortical structural complexity increase across the early life course during maturation of the cerebral cortex and add new data to support an association between FD and cognitive ageing. In terms of the neurobiological basis of cognitive reserve we tentatively concluded that more successful cognitive aging could be linked to greater cortical complexity. 3.5. Childhood socioeconomic status influences adult hippocampal size We investigated in the Aberdeen 1936 birth cohort relationships between socioeconomic status (SES) in childhood and magnetic resonance imaging (MRI)-derived brain volume measures typical of brain aging and Alzheimer's disease (AD). Using a cross-sectional and longitudinal observational approach, we invited volunteers without dementia of whom 249–320 (77%) agreed. We measured whole brain and hippocampal volumes and
68
L.J Whalley et al. / Psychiatry Research: Neuroimaging 247 (2016) 65–70
Fractal Dimension measures the complexity of the edge
South Africa: FD = 1.05 Australia: FD = 1.13 Great Britain: FD = 1.25 Norway: FD = 1.52
Fig. 1
Table 1 The associations between life course cognitive changes and fractal dimension (FD) measures.
Examples of structural comparison between the low and the high fractal of white matter in three different views; coronal, sagittal and axial (the white matter is represented in 2D images for purpose of illustration only). A: subject with low FD; B: subject with high FD. Visual inspection showed that subject with a more complex white matter structure and a more regular surface has respectively a higher FD. Fig. 2
recorded childhood SES history, the number of years of education undertaken, and adult SES history.We used FreeSurfer (http://sur fer.nmr.mgh.harvard.edu) to measure hippocampal volume from T1 weighted MRI brain images (Fig. 3). Analysis of hippocampal volumes showed a significant association between childhood SES and hippocampal volume after adjusting for mental ability at age 11 years, adult SES, gender, and education (Staff et al., 2012): a significant association between childhood SES and hippocampal volumes in late life is consistent with the established neurodevelopmental findings that early life
Model
Early life maturation significance
Fluid ability change
Life course decline
1. Sex WMV FD 2. Sex WMV χ2 3. Sex WMV FD χ2
p ¼0.09 p ¼0.60 p ¼0.11 p ¼0.95
p ¼ 0.001a p ¼ 0.009a p ¼ 0.007a p ¼ 0.13
P ¼0.05a P ¼0.03a P ¼0.20 P ¼0.11
Abbreviations: FD – Fractal Dimension; χ2 ¼ fractal fit, WMV – white matter volume. a A significant association between fractal measures and cognitive change scores.
conditions have an effect on structural brain development. This remains detectable more than 50 years later. The hippocampus is critical in memory formation and is reduced in size as Alzheimer's disease progresses. An association between hippocampal volume and childhood SES in older adults without dementia could reflect effects of relative adversity, fewer educational opportunities, poor nutrition and exposure to greater psychosocial stress. Any of the foregoing could be relevant either alone or in combination. These effects could continue into adult life as the disadvantages of lower SES persist for many individuals. Could a smaller hippocampus contribute to cognitive reserve? If a smaller hippocampus is more susceptible to the effects of brain aging because it is smaller, this would be consistent with the view
L.J Whalley et al. / Psychiatry Research: Neuroimaging 247 (2016) 65–70
69
Fig. 3
that the passive component of reserve is influenced by total or regional brain volume. An association age three through 20 years between parental education or income and volumes of brain structures of their children that are critical to higher cognitive functions was reported by Noble et al. (2015). These findings suggest that the influences of socioeconomic factors on brain aging in late adulthood may have already been exerted during brain development. If supported, this would strengthen the hypothesis that individual differences in rates of brain aging originate during childhood and adolescence (Whalley, 2015: 274–302). 3.6. Is there a balance between cognitive reserve, cerebrovascular disease and Alzheimer's disease? The cognitive reserve hypothesis seeks to explain discrepancy between clinical severity of dementia and the nature and extent of brain pathology. In this sixth and last example from our research programme, we examined the balance between brain magnetic resonance imaging measures of the two most common pathologies associated with brain ageing,(cerebrovascular disease and Alzheimer's disease), with parameters of cerebral reserve in wellcharacterized sample (Murray et al., 2011). These were drawn from participants in the Aberdeen 1936 birth cohort described above and for whom childhood IQ were available. Using brain magnetic resonance imaging, we quantified cerebrovascular disease by measuring brain white matter hyperintensities on fluid attenuation inversion recovery images using Scheltens' scale and Alzheimer's like changes was measured from volumetric data using FreeSurfer (http://surfer.nmr.mgh.harvard.edu) to extract whole brain volume and hippocampal volumes in turn. The effect of these measures of brain burden on life-long cognitive ageing from the age of 11–68 years was compared using structural equation modelling with the effect of educational attainment and occupational grade. Complete brain burden and reserve data were available in 224 participants. We found that educational attainment, but not occupation, has a measurable and positive effect, with a standardized regression weight of þ 0.23, on late life cognitive ability in people without cognitive impairment aged 68 years, allowing for the influence of childhood intelligence and the two most common subclinical brain pathological burdens in the ageing brain. In addition, we demonstrate that the magnitude of the contribution of education is greater than the negative impact of either neuropathological burden alone, with standardized regression weights of 0.14 for white matter hyperintensities and 0.20 for hippocampal atrophy. This study
illustrates how education counteracts the deleterious effects of cerebrovascular disease and Alzheimer's disease and highlights the importance of quantifying cognitive reserve in dementia research. 3.7. Do childhood socioeconomic circumstances influence the occurrence of brain hyperintensities? There are many data that support the foetal origins of late life vascular disorders and diabetes (Whalley et al., 2006). Therefore, we hypothesised that brain hyperintensities which appear more often in the presence of vascular risk factors would also be associated with lower socioeconomic circumstances in childhood. Murray et al. (2014) showed that disadvantaged early life socioeconomic circumstances are associated with increased brain imaging evidence of cerebrovascular disease in late midlife, with its established negative consequences for cognition, stroke, dementia and survival. The link between childhood socioeconomic circumstances and brain hypertintensities was of similar effect size but independent from the effect of hypertension. Together with our data (referenced above, 3.5) which demonstrated smaller hippocampal volumes in those from poorer childhood socioeconomic circumstances these results indicate that childhood disadvantage imposes a double disadvantage of brain burden. They suggest plausible pathways towards greater dementia prevalence in those from disadvantaged socioeconomic backgrounds.
4. Comment Our research programme has enjoyed unique access to archived IQ test scores obtained in historical national surveys of the mental ability of Scottish schoolchildren. With the support of the survivors of those surveys, we have obtained measures of cerebral cortical structure and function and related these to current cognition with and without adjustment for childhood IQ. These analyses have supported the conclusion that the adverse effects of brain aging are mitigated at least in part by both active and passive processes. We understood that education and occupational complexity contributed to active processes and supported this through measurement of their effects in structural equation models of our data. We established the programme on brain aging and health with the aim of identifying factors that would inform the design of intervention studies undertaken to prevent or slow dementia
70
L.J Whalley et al. / Psychiatry Research: Neuroimaging 247 (2016) 65–70
onset. Like many other studies, we showed that education attainments and occupational level contributed positively to cognitive function in late adulthood and that their effects remained after adjustment for childhood IQ. We interpret these effects as evidence in support of the proposal that steps to prevent dementia will acknowledge findings of this type and seek to modify life course factors that might reduce dementia risk (Whalley, 2015: 331–351). Our finding of reduced hippocampal volumes associated with lower socioeconomic status in childhood is consistent with this overall view and, potentially, will prove relevant to the search for biomarkers of increased susceptibility to dementia and cognitive aging, Demonstration of a balance between cognitive reserve and agerelated brain pathologies of cerebrovascular disease and Alzheimer's disease makes intuitive good sense that could contribute to the advice offered in health education. In this context, it is possible to envisage acceptance of advice to reduce risk of cerebrovascular disease through life style changes and blood pressure management. Reduction of Alzheimer risk remains to be established and probably awaits fundamental drug discovery rather than the implementation of findings from epidemiological surveys. Promotion of cognitive reserve remains a life course objective, probably only achievable through targeted educational programmes, a socially engaged and cognitively demanding lifestyle and the judicious consumption of a diet that is restricted in calories but replete in antioxidants, B vitamins and marine oils.
Acknowledgements The Aberdeen birth Cohort Studies were established with grants to Lawrence Whalley by the Henry Smith Charity, the UK Biotechnology and Biological Sciences Research Council and a Professorial Clinical Fellowship Award from the Wellcome Trust. The imaging studies reported here were supported by grants to all three authors by the Chief Scientist Organisation of the Scottish Health Department and Alzheimer Research UK. We are grateful to the volunteers in the Aberdeen 1921 and 1936 Birth Cohort Studies and to our research colleagues in the Aberdeen biomedical Imaging Centre (Drs. Ahearn, Waiter, and Mustafa) and our long-term collaborators in the University of Edinburgh (Professors Deary and Starr at www.ccace.ed.ac.uk).
References Angevaren, M., Aufdemkampe, G., Verhaar, H.J., Aleman, A., Vanhees, L., 2008. Physical activity and enhanced fitness to improve cognitive function in older people without known cognitive impairment. Cochrane Database Syst. Rev. 16 (2), CD005381. Barnes, D.E., Santos-Modesitt, W., Poelke, G., Kramer, A.F., Castro, C., Middleton, L.E., Yaffe, K., 2013. The Mental Activity and eXercise (MAX) trial: a randomized controlled trial to enhance cognitive function in older adults. J. Am. Med. Assoc. Intern. Med. 173, 797–804. Deary, I.J., Simonotto, E., Meyer, M., Marshall, A., Marshall, I., Goddard, N., Wardlaw, J.M., 2004. The functional anatomy of inspection time: an event-related fMRI
study. NeuroImage 22, 1466–1479. Deary, I.J., Whalley, L.J., Starr, J.M., 2009. Looking Down on Human Intelligence: Follow-up Studies of the Scottish Mental Surveys 1932 and 1947. American Psychological Association. Fratiglioni, L., Qiu, C., 2011. Prevention of cognitive decline in ageing: dementia as the target, delayed onset as the goal. Lancet Neurol. 10, 778–779. Gunning-Dixon, F.M., Raz, N., 2000. The cognitive correlates of white matter abnormalities in normal aging: a quantitative review. Neuropsychology 14, 224–232. King, R.D., George, A.T., Jeon, T., Hynan, L.S., Youn, T.S., Kennedy, D.N., Dickerson, B., 2009. Characterization of atrophic changes in the cerebral cortex using fractal dimensional analysis. Brain Imaging Behav. 3, 154–166. Matthews, F.E., Arthur, A., Barnes, L.E., Bond, J., Jagger, C., Robinson, L., Brayne, C., 2013. Medical Research Council Cognitive Function and Ageing Collaboration, 2013. A two-decade comparison of prevalence of dementia in individuals aged 65 years and older from three geographical areas of England: results of the Cognitive Function and Ageing Study I and II. Lancet (epub 16th July). McDaniel, M.A., 2005. Big-brained people are smarter: a meta-analysis of the relationship between in vivo brain volume and intelligence. Intelligence 33, 337–346. Murray, A.D., McNeil, C.J., Salarirad, S., Whalley, L.J., Staff, R.T., 2014. Early life socioeconomic circumstance and late life brain hyperintensities – a population based cohort study. PLoS ONE 9 (2), e88969. Murray, A.D., Staff, R.T., McNeil, C.J., Salarirad, S., Ahearn, T.S., Mustafa, N., Whalley, L.J., 2011. The balance between cognitive reserve and brain imaging biomarkers of cerebrovascular and Alzheimer's diseases. Brain 134, 3687–3696. McGovern., P.G., Pankow, J.S., Shahar, E., Doliszny, K.M., Folsom, M.D., Blackburn, H., m Luepker, R.V., 1996. Recent trends in acute coronary heart disease. N. Engl. J. Med. 334, 884–890. Mustafa, N., Ahearn, T.S., Waiter, G.D., Murray, A.D., Whalley, L.J., Staff, R.T., 2012. Brain structural complexity and life course cognitive change. NeuroImage 61; , pp. 664–701. Noble, K.G., Huston, S.M., Brito, N.H., Bartsch, H., Kan, E., Kupperman, J.M., et al., 2015. Family income, parental education and brains tructure in children and adolescents. Nat. Neurosci. 18, 773–778. http://dx.doi.org/10.1038/nn.3983. Qiu, C., von Strauss, E., Backman., L., Winblad, B., Fratiglioni, L., 2013. Twenty-year changes in dementia occurrence suggest decreasing incidence in central Stockholm, Sweden. Neurology 80, 1888–1894. Shyu, K.K., Wu, Y.T., Chen, T.R., Chen, H.Y., Hu, H.H., Guo, W.Y., 2010. Measuring complexity of fetal cortical surface from MR images using 3-D modified boxcounting method. Inst. Electr. Electron. Eng. Trans. Instrum. Meas. 99, 1–10. Soros, P., Whitehead, S., Spence, J.D., Hachinski, V., 2012. Opinion: antihypertensive treatment can prevent stroke and cognitive decline. Nat. Rev. Neurol. 9, 174–178. Staff, R.T., Murray, A.D., Deary, I.J., Whalley, L.J., 2004. What provides cerebral reserve? Brain 127, 1191–1199. Staff, R.T., Murray, A.D., Deary, I.J., Whalley, L.J., 2006. Generality and specificity in cognitive aging; a volumetric brain analysis. NeuroImage 30, 1433–1440. Staff, R.T., Murray, A.D., Ahearn, T.S., Mustafa, N., Fox, H.C., Whalley, L.J., 2012. Childhood socioeconomic status and adult brain size: childhood socioeconomic status influences adult hippocampal size. Ann. Neurol. 71, 653–660. Stern, Y., 2002. What is cognitive reserve? Theory and research application of the reserve concept. J. Int. Neuropsychol. Soc. 8, 448–460. Takahashi, T., Murata, T., Omori, M., Kosaka, H., Takahashi, K., Yonekura, Y., Wada, Y., 2004. Quantitative evaluation of age-related white matter microstructural changes on MRI by multifractal analysis. J. Neurol. Sci. 225, 33–37. Waiter, G.D., Fox, H.C., Murray, A.D., Starr, J.M., Staff, R.T., Bourne, V.J., Deary, I.J., 2008. Is retaining the youthful functional anatomy underlying speed of information processing a signature of successful cognitive ageing? An event-related fMRI study of inspection time performance. NeuroImage 41, 581–595. Whalley, L.J., Murray, A.D., Staff, R.T., Starr, J.M., Deary, I.J., Fox, H.C., Lemmon, H., Duthie, S.J., Collins, A.R., Crawford, J.R., 2011. How the 1932 and 1947 mental surveys of Aberdeen schoolchildren provide a framework to explore the childhood origins of late onset disease and disability. Maturitas 69, 365–372. Whalley, L.J., Deary, I.J., Appleton, C.L., Starr, J.M., 2004. Cognitive reserve and the neurobiology of cognitive aging. Ageing Res. Rev. 3, 369–382. Whalley, L.J., 2015. Understanding Brain Aging and Dementia: A Life Course Approach. Columbia University Press, New York. Whalley, L.J., Dick, F., McNeil, G., 2006. A life course approach to the aetiology of late onset dementias. Lancet Neurol. 5, 87–96.
Contents lists available at ScienceDirect
Psychiatry Research: Neuroimaging journal homepage: www.elsevier.com/locate/psychresns
Review article
Cerebral correlates of cognitive reserve Lawrence J Whalley a,b,n, Roger T Staff a,c, Helen C Fox a, Alison D Murray a a b c
Aberdeen Biomedical Imaging Centre, College of Life Sciences and Medicine, University of Aberdeen, Aberdeen AB25 2ZD, UK Institute of Applied Health Sciences, University of Aberdeen, Aberdeen, UK NHS Grampian, Foresterhill, Aberdeen, UK
art ic l e i nf o
a b s t r a c t
Article history: Received 9 October 2015 Accepted 13 October 2015 Available online 19 October 2015
Cognitive reserve is a hypothetical concept introduced to explain discrepancies between severity of clinical dementia syndromes and the extent of dementia pathology. We examined cognitive reserve in a research programme that followed up a non-clinical sample born in 1921 or 1936 and IQ-tested age 11 years in 1932 or 1947. Structural MRI exams were acquired in about 50% of the sample from whom a subsample were recruited into an additional fMRI study. Here, we summarise findings from seven interrelated studies. These support an understanding of cognitive reserve as a balance between positive life course activity-driven experiences and the negative effects of brain pathologies including cerebrovascular disease and total and regional brain volume loss. Hypothesised structural equation models illustrate the relative causal effects of these positive and negative contributions. Cognitive reserve is considered in the context of choice of interventions to prevent dementia and the opposing effects of cerebrovascular disease and Alzheimer like brain appearances. Crown Copyright & 2015 Published by Elsevier Ireland Ltd. All rights reserved.
Keywords: Cognitive reserve Childhood IQ White matter hyperintensities Functional MRI Cortical complexity Structural equation model Cognitive aging Alzheimer's disease
Contents 1.
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1. Cerebral correlates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2. Cognitive reserve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. The Aberdeen birth cohort studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. The studies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Cognitive reserve and brain volume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Intelligence, cognitive reserve and brain aging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3. An fMRI study of successful cognitive aging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.4. Cortical complexity and cognitive aging. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.5. Childhood socioeconomic status influences adult hippocampal size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.6. Is there a balance between cognitive reserve, cerebrovascular disease and Alzheimer's disease? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.7. Do childhood socioeconomic circumstances influence the occurrence of brain hyperintensities? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Comment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1. Introduction Population
trends
in
age
structure
accelerated
rapidly
n Corresponding author at: Aberdeen Biomedical Imaging Centre, College of Life Sciences and Medicine, University of Aberdeen, Aberdeen AB25 2ZD, UK. E-mail address: [email protected] (L. Whalley).
65 66 66 66 66 66 66 66 67 67 67 69 69 69 70 70
throughout the 20th century. As numbers of old people increased, so concerns grew that diseases associated with aging would increase in prevalence and potentially jeopardise health care provisions for the elderly. Dementia is a major cause of disability and death affecting about 5% of all those over age 65 years and about 20% of those aged 85 years or more. Although recent reports of falling prevalence of dementia give rise to cautious optimism that
http://dx.doi.org/10.1016/j.pscychresns.2015.10.012 0925-4927/Crown Copyright & 2015 Published by Elsevier Ireland Ltd. All rights reserved.
66
L.J Whalley et al. / Psychiatry Research: Neuroimaging 247 (2016) 65–70
dementia may prove somewhat less of a burden on health services, effective dementia prevention strategies remain a major priority in the 21st century. Disappointed by the results of recent trials of anti-amyloid therapies in Alzheimer's disease (AD), alternative approaches to dementia prevention are now widely discussed. Attention has turned to possible explanations of decreased AD prevalence detected in surveys in Sweden (Qiu et al., 2013) and England (Matthews et al., 2013) and comparable data from Holland and the US. From these reports, two broad themes emerged. The first concerns comparisons between recent reduced dementia prevalence and established downward trends in heart-disease and stroke (McGovern et al., 1996; Soros et al., 2012). The second, draws on reports of the benefits for cognitive aging and possible dementia protection that can be linked to lifestyles that are cognitively demanding, socially engaged and physically active (Angevaren et al., 2008; Fratiglioni and Qiu, 2011; Barnes et al., 2013). In this review, we describe our approach to the concept of cognitive reserve. Our aim is to make tractable the investigation of cognitive reserve with the intention of identifying those components that could enhance its protective or compensatory potential and inform the design of studies to prevent or delay dementia onset.
2. Methods 2.1. The Aberdeen birth cohort studies Since 1998, we have followed up local community residents who at age about 11 years were at school in Aberdeen and sat a group-administered IQ-type test. Full details are given elsewhere (Whalley et al., 2011). Briefly, the Aberdeen birth cohort studies were started in 1998 when Scottish Mental survey archives of the Scottish Council for Research in Education were re-discovered and permissions obtained to follow-up survivors born in 1921 or 1936 and then aged about 77 or 64 years and who had entered (or were about to enter) the age of greatest risk for Alzheimer's disease (AD). Sources of attrition from the study, exposures to childhood adversity, nutritional, genetic and life style factors of possible relevance to extent of age-related cognitive decline and the timing of onset of dementia are given elsewhere (Whalley et al. 2011). By 2014, the feasibility of following up more than 75% of Scottish Mental Survey survivors living in the Aberdeen area without dementia was well-established, dementia ascertainment to age about 88 years was completed in the 1921 birth cohort and was underway in the 1936 born cohort (by then aged about 78 years). Structural MRI exams were introduced in 2001 and, by 2008, about 50% of the total sample of 791 participants has undergone at least one MRI. These databases are available to other bone fide research groups wishing to test specific hypotheses that may either replicate their own findings or make best use of the data collected in the Aberdeen studies (contact [email protected]).
3. The studies 3.1. Cognitive reserve and brain volume
1.1. Cerebral correlates The cerebral cortex is a complex layered and much folded bilateral structure that makes up the cerebral hemispheres. Agerelated changes in the cerebral cortex are multi-factorial and affect neurones and glia, the cerebral vasculature and the white matter tracts. In life, these changes are detectable using advanced brain imaging techniques that accurately measure brain structures, vascular pathologies (including small vessel disease) and regional brain metabolic differences. Our studies on cognitive reserve have largely relied on structural magnetic resonance imaging (MRI) and, to a much lesser extent, functional MRI (fMRI) to reveal differences in blood oxygen level detection (BOLD) responses in comparisons between experimental conditions. 1.2. Cognitive reserve Quantification of cognitive reserve depends upon the precise measurement of cognitive performance. A wide range of cognitive domains makes up overall cognitive function and shares common variance (a general factor ‘g’) that explains the extent to which domains are inter-correlated. In our studies we have selected a small number of standardised cognitive tests that capture verbal learning (auditory verbal learning test, AVLT) and spatial ability (Block Design, BD); non-verbal reasoning, Raven's Standardised Progressive Matrices (RPM); mental speed (Digit Symbol, DS) and verbal fluency (Guilford's Uses of Common Objects). Baseline childhood IQ was taken from historical records of testing of children using the Moray House Test (Deary et al., 2009). Age-related cognitive change was estimated from the difference between observed current cognitive performance in late adulthood and cognitive performance predicted from Moray House Test scores converted to standard IQ-type scores (mean 100, standard deviation 15). It was hypothesised that cognitive reserve would be a major positive influence on individual differences between observed current cognitive performance and expected performance predicted by childhood IQ. Negative influences include age-related brain pathology (brain atrophy and brain lesion load thought to be of cerebrovascular origin).
In our first investigation into the concept of cognitive reserve, we explored the possibility that reserve was influenced by head size (as a proxy for lifetime maximum brain size), duration of education, and occupational attainment in 92 volunteers born in 1921 and whose childhood IQ age 11 years was available. We performed neuropsychological tests of verbal memory, spatial ability, non-verbal abstract reasoning, verbal fluency and mental speed at age about 79 years and within 3 months conducted MRI exams. The fraction of total intracranial volume occupied by brain at age 79 years was used as a measure of brain shrinkage (atrophy) and density of white matter hyperintensities as a measure of cerebrovascular disease (burden). The results showed that education and occupational attainment but not total intracranial volume (a proxy for maximum brain size) contribute in this preliminary study to retention of cognitive performance from age 11 to 79 years (Staff et al., 2004). Because we had baseline childhood IQ scores, we could adjust age 79 cognitive test scores for baseline and use general linear modelling of covariance to test cognitive reserve hypotheses. Separate models were tested for verbal memory and non-verbal reasoning. The simplest model of cerebral reserve is that it is a passive property that protects the individual from age-related cognitive decline. This line of reasoning implies that a person with a larger brain is capable of withstanding a greater degree of age-related pathology and would decline less than a person with a smaller brain. An active model of cognitive reserve can be tested in a similar way after adding education as a likely contributor to verbal memory but not to non-verbal reasoning. These models of active and passive reserve were first suggested by Stern (2002). 3.2. Intelligence, cognitive reserve and brain aging In a later analysis, using the same MRI data, and after accounting for childhood IQ, we explored relationships between specific cognitive domains, the general factor (‘g’) that explained variance shared between test and brain volumes. We hypothesised that brain aging would be correlated with non-uniform reductions in brain volumes and that these would be associated with scores on specific cognitive tests (Staff et al., 2006). As predicted, we found links between specific cognitive tests and brain volumes. An
L.J Whalley et al. / Psychiatry Research: Neuroimaging 247 (2016) 65–70
association was also found between brain volumes and the general factor ‘g’. After removing the influence of ‘g’ from each of the specific cognitive test scores, no remaining significant associations were found between brain volumes and specific cognitive domain test scores. We concluded that cognitive aging across multiple domains of cognitive function is associated with brain volume differences but that – in this small sample of older adults without dementia – we did not find evidence that specific components of cognitive aging were associated with smaller brain volumes. We found preliminary evidence that the white matter tracts of the left (dominant) hemisphere were reduced in size and that this correlated with overall cognitive aging. We tentatively concluded that these reductions were causally linked to cognitive aging but since ours was a cross-sectional and not a longitudinal study, we cannot be sure that these did not pre-date the onset of cognitive aging. An alternative explanation would be that these white matter tracts reflect a pre-existing “passive” component of cognitive reserve that allows an individual to adjust more effectively to brain aging. Other studies have found these white tracts are directly linked in size to chronological age and this would support their causal rather than a reserve role in cognitive aging (Gunning-Dixon and Raz, 2000). 3.3. An fMRI study of successful cognitive aging The Aberdeen 1936 birth cohort was recruited aged about 64 years. At age 70 years, a small sub-sample participated in an fMRI study. At this stage in the development of our research programme, we were interested in neurobiological mechanisms that could underpin (i.e. be “fundamental”) to differences in rates of cognitive aging. We anticipated that at least in part these biological findings would prove relevant to understanding the neurobiology of cognitive reserve (Whalley et al., 2004). Briefly, the design of the experiment was as follows. We compared two groups of non-demented 70-year-olds who, at age 11, had relatively similar general cognitive ability but who, in old age, had diverged, with one group demonstrating relatively successful (“sustainers” N ¼25), and the other one unsuccessful, cognitive ageing (“decliners” N ¼15). We examined whether, in older people with relatively successful cognitive ageing, their BOLD activation– deactivation pattern while they performed an inspection time task was more or less similar to those of healthy younger individuals than older people with relatively unsuccessful cognitive ageing (Waiter et al., 2008). Inspection time is a two-alternative forced choice, backward masking test of the speed of the early stages of visual information processing. Inspection time has a well-established, significant association with higher cognitive abilities. The group of cognitive sustainers showed a pattern of BOLD activation–deactivation in response to inspection time stimulus duration differences that was similar to a healthy young sample (Deary et al., 2004). The group of cognitive decliners lacked these clear neural networks. The relative preservation of complex reasoning skills in old age may be associated with the preservation of the neural networks that underpin information processing in youth. We concluded that retention of youthful neural networks by this small group of “sustainers” who had aged relatively successfully to that point provided a neurobiological basis for their retention of complex reasoning skills. These advantages may be generalisable to other cognitive domains and prove relevant to understanding the neurobiological basis of cognitive reserve. Age-related cognitive performance does not invariably decline and some individuals will improve in late adulthood. One explanation, with strong empirical support, is that the slowing of information processing speed is an important influence on cognitive aging. The neurobiology of age-related changes in processing speed is poorly understood. Sampling successful agers using a
67
strategy based on complex tests of reasoning, showed brain patterns of activity that resemble younger people's during a visual information processing task. 3.4. Cortical complexity and cognitive aging We recognised that in order to understand how neural networks performed in supporting cognition in the face of brain aging, our next step should aim to quantify structural differences in the complexity of neural networks. We hypothesised that successful cognitive agers would have more complex cerebral cortices than individuals who declined. Previously, we had rejected the “more is better” hypothesis of successful cognitive aging so began to explore alternative approaches to cortical structure. McDaniel (2005) had presented a meta-analysis of reports on brain size and general mental ability and concluded that brain size made a modest but significantcontribution to IQ. The cerebral cortex is a fractal like structure made of repeated elements each of which resembles the structure as a whole. Cortical fractal structure can be characterised by a single numerical value – the fractal dimension (FD) – that summarises the irregularity and reproducibility at different scales of (1) the external cortical surface and (2) the internal boundary between sub-cortical grey and white matter. Figs. 1 and 2 illustrate how FD increases with increasing complexity of the surface under investigation. This is easily pictured in comparisons between coastlines (Fig. 1) though less readily seen in cortical surfaces (Fig. 2). In the development and aging of the human brain, there is increasing cortical complexity from early foetal life (Shyu et al., 2010) and into adulthood (Takahashi et al. 2004) until in late life when decreasing cortical complexity is seen (King et al., 2009). We decided to examine associations between the FD of white matter and cognitive changes across the life course in the absence of detectable brain disease (Mustafa et al., 2012). The FD was calculated from segmented cerebral white matter MR images in 217 subjects aged about 68 years. They were sampled from the Aberdeen 1936 birth cohort described above for whom we had archived IQ scores from age 11 years. Cognitive test scores of fluid and crystallised intelligence were obtained at the time of MR imaging. Significant differences were found (in intracranial volume, brain volume, white matter volume and Raven's Progressive Matrices score) between men and women at age 68 years and novel associations were found between FD and measures of cognitive change over the life course from age 11 to 68 years. Those with greater FD were found to have greater than expected fluid abilities at age 68 years than predicted by their childhood intelligence and less cognitive decline from age 11 to 68 years (Table 1). These results are consistent with other reports that FD measures of cortical structural complexity increase across the early life course during maturation of the cerebral cortex and add new data to support an association between FD and cognitive ageing. In terms of the neurobiological basis of cognitive reserve we tentatively concluded that more successful cognitive aging could be linked to greater cortical complexity. 3.5. Childhood socioeconomic status influences adult hippocampal size We investigated in the Aberdeen 1936 birth cohort relationships between socioeconomic status (SES) in childhood and magnetic resonance imaging (MRI)-derived brain volume measures typical of brain aging and Alzheimer's disease (AD). Using a cross-sectional and longitudinal observational approach, we invited volunteers without dementia of whom 249–320 (77%) agreed. We measured whole brain and hippocampal volumes and
68
L.J Whalley et al. / Psychiatry Research: Neuroimaging 247 (2016) 65–70
Fractal Dimension measures the complexity of the edge
South Africa: FD = 1.05 Australia: FD = 1.13 Great Britain: FD = 1.25 Norway: FD = 1.52
Fig. 1
Table 1 The associations between life course cognitive changes and fractal dimension (FD) measures.
Examples of structural comparison between the low and the high fractal of white matter in three different views; coronal, sagittal and axial (the white matter is represented in 2D images for purpose of illustration only). A: subject with low FD; B: subject with high FD. Visual inspection showed that subject with a more complex white matter structure and a more regular surface has respectively a higher FD. Fig. 2
recorded childhood SES history, the number of years of education undertaken, and adult SES history.We used FreeSurfer (http://sur fer.nmr.mgh.harvard.edu) to measure hippocampal volume from T1 weighted MRI brain images (Fig. 3). Analysis of hippocampal volumes showed a significant association between childhood SES and hippocampal volume after adjusting for mental ability at age 11 years, adult SES, gender, and education (Staff et al., 2012): a significant association between childhood SES and hippocampal volumes in late life is consistent with the established neurodevelopmental findings that early life
Model
Early life maturation significance
Fluid ability change
Life course decline
1. Sex WMV FD 2. Sex WMV χ2 3. Sex WMV FD χ2
p ¼0.09 p ¼0.60 p ¼0.11 p ¼0.95
p ¼ 0.001a p ¼ 0.009a p ¼ 0.007a p ¼ 0.13
P ¼0.05a P ¼0.03a P ¼0.20 P ¼0.11
Abbreviations: FD – Fractal Dimension; χ2 ¼ fractal fit, WMV – white matter volume. a A significant association between fractal measures and cognitive change scores.
conditions have an effect on structural brain development. This remains detectable more than 50 years later. The hippocampus is critical in memory formation and is reduced in size as Alzheimer's disease progresses. An association between hippocampal volume and childhood SES in older adults without dementia could reflect effects of relative adversity, fewer educational opportunities, poor nutrition and exposure to greater psychosocial stress. Any of the foregoing could be relevant either alone or in combination. These effects could continue into adult life as the disadvantages of lower SES persist for many individuals. Could a smaller hippocampus contribute to cognitive reserve? If a smaller hippocampus is more susceptible to the effects of brain aging because it is smaller, this would be consistent with the view
L.J Whalley et al. / Psychiatry Research: Neuroimaging 247 (2016) 65–70
69
Fig. 3
that the passive component of reserve is influenced by total or regional brain volume. An association age three through 20 years between parental education or income and volumes of brain structures of their children that are critical to higher cognitive functions was reported by Noble et al. (2015). These findings suggest that the influences of socioeconomic factors on brain aging in late adulthood may have already been exerted during brain development. If supported, this would strengthen the hypothesis that individual differences in rates of brain aging originate during childhood and adolescence (Whalley, 2015: 274–302). 3.6. Is there a balance between cognitive reserve, cerebrovascular disease and Alzheimer's disease? The cognitive reserve hypothesis seeks to explain discrepancy between clinical severity of dementia and the nature and extent of brain pathology. In this sixth and last example from our research programme, we examined the balance between brain magnetic resonance imaging measures of the two most common pathologies associated with brain ageing,(cerebrovascular disease and Alzheimer's disease), with parameters of cerebral reserve in wellcharacterized sample (Murray et al., 2011). These were drawn from participants in the Aberdeen 1936 birth cohort described above and for whom childhood IQ were available. Using brain magnetic resonance imaging, we quantified cerebrovascular disease by measuring brain white matter hyperintensities on fluid attenuation inversion recovery images using Scheltens' scale and Alzheimer's like changes was measured from volumetric data using FreeSurfer (http://surfer.nmr.mgh.harvard.edu) to extract whole brain volume and hippocampal volumes in turn. The effect of these measures of brain burden on life-long cognitive ageing from the age of 11–68 years was compared using structural equation modelling with the effect of educational attainment and occupational grade. Complete brain burden and reserve data were available in 224 participants. We found that educational attainment, but not occupation, has a measurable and positive effect, with a standardized regression weight of þ 0.23, on late life cognitive ability in people without cognitive impairment aged 68 years, allowing for the influence of childhood intelligence and the two most common subclinical brain pathological burdens in the ageing brain. In addition, we demonstrate that the magnitude of the contribution of education is greater than the negative impact of either neuropathological burden alone, with standardized regression weights of 0.14 for white matter hyperintensities and 0.20 for hippocampal atrophy. This study
illustrates how education counteracts the deleterious effects of cerebrovascular disease and Alzheimer's disease and highlights the importance of quantifying cognitive reserve in dementia research. 3.7. Do childhood socioeconomic circumstances influence the occurrence of brain hyperintensities? There are many data that support the foetal origins of late life vascular disorders and diabetes (Whalley et al., 2006). Therefore, we hypothesised that brain hyperintensities which appear more often in the presence of vascular risk factors would also be associated with lower socioeconomic circumstances in childhood. Murray et al. (2014) showed that disadvantaged early life socioeconomic circumstances are associated with increased brain imaging evidence of cerebrovascular disease in late midlife, with its established negative consequences for cognition, stroke, dementia and survival. The link between childhood socioeconomic circumstances and brain hypertintensities was of similar effect size but independent from the effect of hypertension. Together with our data (referenced above, 3.5) which demonstrated smaller hippocampal volumes in those from poorer childhood socioeconomic circumstances these results indicate that childhood disadvantage imposes a double disadvantage of brain burden. They suggest plausible pathways towards greater dementia prevalence in those from disadvantaged socioeconomic backgrounds.
4. Comment Our research programme has enjoyed unique access to archived IQ test scores obtained in historical national surveys of the mental ability of Scottish schoolchildren. With the support of the survivors of those surveys, we have obtained measures of cerebral cortical structure and function and related these to current cognition with and without adjustment for childhood IQ. These analyses have supported the conclusion that the adverse effects of brain aging are mitigated at least in part by both active and passive processes. We understood that education and occupational complexity contributed to active processes and supported this through measurement of their effects in structural equation models of our data. We established the programme on brain aging and health with the aim of identifying factors that would inform the design of intervention studies undertaken to prevent or slow dementia
70
L.J Whalley et al. / Psychiatry Research: Neuroimaging 247 (2016) 65–70
onset. Like many other studies, we showed that education attainments and occupational level contributed positively to cognitive function in late adulthood and that their effects remained after adjustment for childhood IQ. We interpret these effects as evidence in support of the proposal that steps to prevent dementia will acknowledge findings of this type and seek to modify life course factors that might reduce dementia risk (Whalley, 2015: 331–351). Our finding of reduced hippocampal volumes associated with lower socioeconomic status in childhood is consistent with this overall view and, potentially, will prove relevant to the search for biomarkers of increased susceptibility to dementia and cognitive aging, Demonstration of a balance between cognitive reserve and agerelated brain pathologies of cerebrovascular disease and Alzheimer's disease makes intuitive good sense that could contribute to the advice offered in health education. In this context, it is possible to envisage acceptance of advice to reduce risk of cerebrovascular disease through life style changes and blood pressure management. Reduction of Alzheimer risk remains to be established and probably awaits fundamental drug discovery rather than the implementation of findings from epidemiological surveys. Promotion of cognitive reserve remains a life course objective, probably only achievable through targeted educational programmes, a socially engaged and cognitively demanding lifestyle and the judicious consumption of a diet that is restricted in calories but replete in antioxidants, B vitamins and marine oils.
Acknowledgements The Aberdeen birth Cohort Studies were established with grants to Lawrence Whalley by the Henry Smith Charity, the UK Biotechnology and Biological Sciences Research Council and a Professorial Clinical Fellowship Award from the Wellcome Trust. The imaging studies reported here were supported by grants to all three authors by the Chief Scientist Organisation of the Scottish Health Department and Alzheimer Research UK. We are grateful to the volunteers in the Aberdeen 1921 and 1936 Birth Cohort Studies and to our research colleagues in the Aberdeen biomedical Imaging Centre (Drs. Ahearn, Waiter, and Mustafa) and our long-term collaborators in the University of Edinburgh (Professors Deary and Starr at www.ccace.ed.ac.uk).
References Angevaren, M., Aufdemkampe, G., Verhaar, H.J., Aleman, A., Vanhees, L., 2008. Physical activity and enhanced fitness to improve cognitive function in older people without known cognitive impairment. Cochrane Database Syst. Rev. 16 (2), CD005381. Barnes, D.E., Santos-Modesitt, W., Poelke, G., Kramer, A.F., Castro, C., Middleton, L.E., Yaffe, K., 2013. The Mental Activity and eXercise (MAX) trial: a randomized controlled trial to enhance cognitive function in older adults. J. Am. Med. Assoc. Intern. Med. 173, 797–804. Deary, I.J., Simonotto, E., Meyer, M., Marshall, A., Marshall, I., Goddard, N., Wardlaw, J.M., 2004. The functional anatomy of inspection time: an event-related fMRI
study. NeuroImage 22, 1466–1479. Deary, I.J., Whalley, L.J., Starr, J.M., 2009. Looking Down on Human Intelligence: Follow-up Studies of the Scottish Mental Surveys 1932 and 1947. American Psychological Association. Fratiglioni, L., Qiu, C., 2011. Prevention of cognitive decline in ageing: dementia as the target, delayed onset as the goal. Lancet Neurol. 10, 778–779. Gunning-Dixon, F.M., Raz, N., 2000. The cognitive correlates of white matter abnormalities in normal aging: a quantitative review. Neuropsychology 14, 224–232. King, R.D., George, A.T., Jeon, T., Hynan, L.S., Youn, T.S., Kennedy, D.N., Dickerson, B., 2009. Characterization of atrophic changes in the cerebral cortex using fractal dimensional analysis. Brain Imaging Behav. 3, 154–166. Matthews, F.E., Arthur, A., Barnes, L.E., Bond, J., Jagger, C., Robinson, L., Brayne, C., 2013. Medical Research Council Cognitive Function and Ageing Collaboration, 2013. A two-decade comparison of prevalence of dementia in individuals aged 65 years and older from three geographical areas of England: results of the Cognitive Function and Ageing Study I and II. Lancet (epub 16th July). McDaniel, M.A., 2005. Big-brained people are smarter: a meta-analysis of the relationship between in vivo brain volume and intelligence. Intelligence 33, 337–346. Murray, A.D., McNeil, C.J., Salarirad, S., Whalley, L.J., Staff, R.T., 2014. Early life socioeconomic circumstance and late life brain hyperintensities – a population based cohort study. PLoS ONE 9 (2), e88969. Murray, A.D., Staff, R.T., McNeil, C.J., Salarirad, S., Ahearn, T.S., Mustafa, N., Whalley, L.J., 2011. The balance between cognitive reserve and brain imaging biomarkers of cerebrovascular and Alzheimer's diseases. Brain 134, 3687–3696. McGovern., P.G., Pankow, J.S., Shahar, E., Doliszny, K.M., Folsom, M.D., Blackburn, H., m Luepker, R.V., 1996. Recent trends in acute coronary heart disease. N. Engl. J. Med. 334, 884–890. Mustafa, N., Ahearn, T.S., Waiter, G.D., Murray, A.D., Whalley, L.J., Staff, R.T., 2012. Brain structural complexity and life course cognitive change. NeuroImage 61; , pp. 664–701. Noble, K.G., Huston, S.M., Brito, N.H., Bartsch, H., Kan, E., Kupperman, J.M., et al., 2015. Family income, parental education and brains tructure in children and adolescents. Nat. Neurosci. 18, 773–778. http://dx.doi.org/10.1038/nn.3983. Qiu, C., von Strauss, E., Backman., L., Winblad, B., Fratiglioni, L., 2013. Twenty-year changes in dementia occurrence suggest decreasing incidence in central Stockholm, Sweden. Neurology 80, 1888–1894. Shyu, K.K., Wu, Y.T., Chen, T.R., Chen, H.Y., Hu, H.H., Guo, W.Y., 2010. Measuring complexity of fetal cortical surface from MR images using 3-D modified boxcounting method. Inst. Electr. Electron. Eng. Trans. Instrum. Meas. 99, 1–10. Soros, P., Whitehead, S., Spence, J.D., Hachinski, V., 2012. Opinion: antihypertensive treatment can prevent stroke and cognitive decline. Nat. Rev. Neurol. 9, 174–178. Staff, R.T., Murray, A.D., Deary, I.J., Whalley, L.J., 2004. What provides cerebral reserve? Brain 127, 1191–1199. Staff, R.T., Murray, A.D., Deary, I.J., Whalley, L.J., 2006. Generality and specificity in cognitive aging; a volumetric brain analysis. NeuroImage 30, 1433–1440. Staff, R.T., Murray, A.D., Ahearn, T.S., Mustafa, N., Fox, H.C., Whalley, L.J., 2012. Childhood socioeconomic status and adult brain size: childhood socioeconomic status influences adult hippocampal size. Ann. Neurol. 71, 653–660. Stern, Y., 2002. What is cognitive reserve? Theory and research application of the reserve concept. J. Int. Neuropsychol. Soc. 8, 448–460. Takahashi, T., Murata, T., Omori, M., Kosaka, H., Takahashi, K., Yonekura, Y., Wada, Y., 2004. Quantitative evaluation of age-related white matter microstructural changes on MRI by multifractal analysis. J. Neurol. Sci. 225, 33–37. Waiter, G.D., Fox, H.C., Murray, A.D., Starr, J.M., Staff, R.T., Bourne, V.J., Deary, I.J., 2008. Is retaining the youthful functional anatomy underlying speed of information processing a signature of successful cognitive ageing? An event-related fMRI study of inspection time performance. NeuroImage 41, 581–595. Whalley, L.J., Murray, A.D., Staff, R.T., Starr, J.M., Deary, I.J., Fox, H.C., Lemmon, H., Duthie, S.J., Collins, A.R., Crawford, J.R., 2011. How the 1932 and 1947 mental surveys of Aberdeen schoolchildren provide a framework to explore the childhood origins of late onset disease and disability. Maturitas 69, 365–372. Whalley, L.J., Deary, I.J., Appleton, C.L., Starr, J.M., 2004. Cognitive reserve and the neurobiology of cognitive aging. Ageing Res. Rev. 3, 369–382. Whalley, L.J., 2015. Understanding Brain Aging and Dementia: A Life Course Approach. Columbia University Press, New York. Whalley, L.J., Dick, F., McNeil, G., 2006. A life course approach to the aetiology of late onset dementias. Lancet Neurol. 5, 87–96.