Data collected from clinical trials and cohort studies, such as dementia studies, are often high-dimensional, censored, heterogeneous and contain missing information, presenting challenges to ...traditional statistical analysis. There is an urgent need for methods that can overcome these challenges to model this complex data. At present there is no cure for dementia and no treatment that can successfully change the course of the disease. Machine learning models that can predict the time until a patient develops dementia are important tools in helping understand dementia risks and can give more accurate results than traditional statistical methods when modelling high-dimensional, heterogeneous, clinical data. This work compares the performance and stability of ten machine learning algorithms, combined with eight feature selection methods, capable of performing survival analysis of high-dimensional, heterogeneous, clinical data. We developed models that predict survival to dementia using baseline data from two different studies. The Sydney Memory and Ageing Study (MAS) is a longitudinal cohort study of 1037 participants, aged 70-90 years, that aims to determine the effects of ageing on cognition. The Alzheimer's Disease Neuroimaging Initiative (ADNI) is a longitudinal study aimed at identifying biomarkers for the early detection and tracking of Alzheimer's disease. Using the concordance index as a measure of performance, our models achieve maximum performance values of 0.82 for MAS and 0.93 For ADNI.
Mild cognitive impairment (MCI) is associated with an increased risk of developing dementia. However, many individuals diagnosed with MCI are found to have reverted to normal cognition on follow-up. ...This study investigated factors predicting or associated with reversion from MCI to normal cognition.
Our analyses considered 223 participants (48.9% male) aged 71-89 years, drawn from the prospective, population-based Sydney Memory and Ageing Study. All were diagnosed with MCI at baseline and subsequently classified with either normal cognition or repeat diagnosis of MCI after two years (a further 11 participants who progressed from MCI to dementia were excluded). Associations with reversion were investigated for (1) baseline factors that included diagnostic features, personality, neuroimaging, sociodemographics, lifestyle, and physical and mental health; (2) longitudinal change in potentially modifiable factors.
There were 66 reverters to normal cognition and 157 non-reverters (stable MCI). Regression analyses identified diagnostic features as most predictive of prognosis, with reversion less likely in participants with multiple-domain MCI (p = 0.011), a moderately or severely impaired cognitive domain (p = 0.002 and p = 0.006), or an informant-based memory complaint (p = 0.031). Reversion was also less likely for participants with arthritis (p = 0.037), but more likely for participants with higher complex mental activity (p = 0.003), greater openness to experience (p = 0.041), better vision (p = 0.014), better smelling ability (p = 0.040), or larger combined volume of the left hippocampus and left amygdala (p<0.040). Reversion was also associated with a larger drop in diastolic blood pressure between baseline and follow-up (p = 0.026).
Numerous factors are associated with reversion from MCI to normal cognition. Assessing these factors could facilitate more accurate prognosis of individuals with MCI. Participation in cognitively enriching activities and efforts to lower blood pressure might promote reversion.
Subjective Cognitive Complaints (SCCs) may represent one of the earliest stages of preclinical dementia. The objective of the present study was to extend previous work by our group to examine the ...relationship between participant-reported and informant-reported memory and non-memory SCCs, cognitive decline and incident dementia, over a six-year period. Participants were 873 community dwelling older adults (Mage = 78.65, SD = 4.79) without dementia and 843 informants (close friends or family) from the Sydney Memory and Ageing Study. Comprehensive neuropsychological testing and diagnostic assessments were carried out at baseline and biennially for six years. Linear mixed models and Cox proportional hazard models were performed to determine the association of SCCs, rate of cognitive decline and risk of incident dementia, controlling demographics and covariates of mood and personality. Participant and informant memory-specific SCCs were associated with rate of global cognitive decline; for individual cognitive domains, participant memory SCCs predicted decline for language, while informant memory SCCs predicted decline for executive function and memory. Odds of incident dementia were associated with baseline participant memory SCCs and informant memory and non-memory SCCs in partially adjusted models. In fully adjusted models, only informant SCCs were associated with increased risk of incident dementia. Self-reported memory-specific cognitive complaints are associated with decline in global cognition over 6-years and may be predictive of incident dementia, particularly if the individual is depressed or anxious and has increased neuroticism or decreased openness. Further, if and where possible, informants should be sought and asked to report on their perceptions of the individual's memory ability and any memory-specific changes that they have noticed as these increase the index of diagnostic suspicion.
Type 2 diabetes (diabetes) is characterized by accelerated cognitive decline and higher dementia risk. Controversy exists regarding the impact of metformin, which is associated with both increased ...and decreased dementia rates. The objective of this study was to determine the association of metformin use with incident dementia and cognitive decline over 6 years in participants with diabetes compared with those not receiving metformin and those without diabetes.
A prospective observational study was conducted of
= 1,037 community-dwelling older participants without dementia aged 70-90 years at baseline (the Sydney Memory and Ageing Study). Exclusion criteria were dementia, major neurological or psychiatric disease, or progressive malignancy. Neuropsychological testing measured cognitive function every 2 years; a battery of tests measured executive function, memory, attention/speed, language, and visuospatial function individually. These were used to determine the measure of global cognition. Incident dementia was ascertained by a multidisciplinary panel. Total brain, hippocampal, and parahippocampal volumes were measured by MRI at baseline and 2 years (
= 526). Data were analyzed by linear mixed modeling, including the covariates of age, sex, education, BMI, heart disease, hypertension, stroke, smoking, and apolipoprotein Eε4 carriage.
Of
= 1,037, 123 had diabetes; 67 received metformin (DM+MF) and were demographically similar to those who did not (DM-noMF) and participants without diabetes (no-DM). DM+MF had significantly slower global cognition and executive function decline compared with DM-noMF. Incident dementia was significantly higher in DM-noMF compared with DM+MF (odds ratio 5.29 95% CI 1.17-23.88;
= 0.05).
Older people with diabetes receiving metformin have slower cognitive decline and lower dementia risk. Large randomized studies in people with and without diabetes will determine whether these associations can be attributed to metformin.
Apolipoproteins have recently been implicated in the etiology of Alzheimer's disease (AD). In particular, Apolipoprotein J (ApoJ or clusterin) has been proposed as a biomarker of the disease at the ...pre-dementia stage. We examined a group of apolipoproteins, including ApoA1, ApoA2, ApoB, ApoC3, ApoE, ApoH and ApoJ, in the plasma of a longitudinal community based cohort.
664 subjects (257 with Mild Cognitive Impairment MCI and 407 with normal cognition), mean age 78 years, from the Sydney Memory and Aging Study (MAS) were followed up over two years. Plasma apolipoprotein levels at baseline (Wave 1) were measured using a multiplex bead fluorescence immunoassay technique.
At Wave 1, MCI subjects had lower levels of ApoA1, ApoA2 and ApoH, and higher levels of ApoE and ApoJ, and a higher ApoB/ApoA1 ratio. Carriers of the apolipoprotein E ε4 allele had significantly lower levels of plasma ApoE, ApoC3 and ApoH and a significantly higher level of ApoB. Global cognitive scores were correlated positively with ApoH and negatively with ApoJ levels. ApoJ and ApoE levels were correlated negatively with grey matter volume and positively with cerebrospinal fluid (CSF) volume on MRI. Lower ApoA1, ApoA2 and ApoH levels, and higher ApoB/ApoA1 ratio, increased the risk of cognitive decline over two years in cognitively normal individuals. ApoA1 was the most significant predictor of decline. These associations remained after statistically controlling for lipid profile. Higher ApoJ levels predicted white matter atrophy over two years.
Elderly individuals with MCI have abnormal apolipoprotein levels, which are related to cognitive function and volumetric MRI measures cross-sectionally and are predictive of cognitive impairment in cognitively normal subjects. ApoA1, ApoH and ApoJ are potential plasma biomarkers of cognitive decline in non-demented elderly individuals.
Alzheimer's disease (AD) is generally considered to be characterized by pathology in gray matter of the brain, but convergent evidence suggests that white matter degradation also plays a vital role ...in its pathogenesis. The evolution of white matter deterioration and its relationship with gray matter atrophy remains elusive in amnestic mild cognitive impairment (aMCI), a prodromal stage of AD.
We studied 155 cognitively normal (CN) and 27 'late' aMCI individuals with stable diagnosis over 2 years, and 39 'early' aMCI individuals who had converted from CN to aMCI at 2-year follow up. Diffusion tensor imaging (DTI) tractography was used to reconstruct six white matter tracts three limbic tracts critical for episodic memory function - the fornix, the parahippocampal cingulum, and the uncinate fasciculus; two cortico-cortical association fiber tracts - superior longitudinal fasciculus and inferior longitudinal fasciculus; and one projection fiber tract - corticospinal tract. Microstructural integrity as measured by fractional anisotropy (FA), mean diffusivity (MD), radial diffusivity (RD) and axial diffusivity (AxD) was assessed for these tracts.
Compared with CN, late aMCI had lower white matter integrity in the fornix, the parahippocampal cingulum, and the uncinate fasciculus, while early aMCI showed white matter damage in the fornix. In addition, fornical measures were correlated with hippocampal atrophy in late aMCI, whereas abnormality of the fornix in early aMCI occurred in the absence of hippocampal atrophy and did not correlate with hippocampal volumes.
Limbic white matter tracts are preferentially affected in the early stages of cognitive dysfunction. Microstructural degradation of the fornix preceding hippocampal atrophy may serve as a novel imaging marker for aMCI at an early stage.
: Incidence of falls in people with cognitive impairment with or without a formal diagnosis of dementia is estimated to be twice that of cognitively intact older adults. This study aimed to ...investigate whether mild cognitive impairment (MCI) is associated with falls in older people.
: Prospective cohort study.
: Community sample, Sydney Memory and Ageing Study.
: A total of 419 nondemented community-dwelling adults, age 70-90 years.
: A comprehensive neuropsychological test battery measuring four cognitive domains provided classification being with or without MCI on the basis of objective published criteria. Assessments of medical, physiologic, and psychological measures were also performed. Fallers were defined as people who had at least one injurious fall or at least two noninjurious falls during a 12-month follow-up period.
: Of the participants, 342 (81.6%) had normal cognitive functioning, 58 (13.8%) had nonamnestic MCI, and 19 (4.5%) had amnestic MCI. People with MCI performed worse than people without MCI in measures of general health and balance. Logistic regression analyses showed that fall risk was significantly greater in people with MCI (odds ratio OR: 1.72, 95% confidence interval 95% CI: 1.03-2.89). This association was mainly apparent when the analysis was restricted to those with nonamnestic MCI (OR: 1.98, 95% CI: 1.11-3.53), where the relationship was primarily explained by impaired executive functioning (OR: 1.27, 95% CI: 1.02-1.59).
: The findings indicate that objectively defined MCI is an independent risk factor for injurious or multiple falls in a representative sample of community-dwelling older people. The presence of nonamnestic MCI, based primarily on executive function, was found to be an important factor in increasing fall risk.
We sought to understand the trajectory of mild cognitive impairment (MCI) better by examining longitudinally different MCI subtypes and progression to dementia and reversion to normal cognition in a ...community sample.
We evaluated the stability of MCI subtypes and risk of dementia over 4 biennial assessments as part of an ongoing prospective cohort study, the Sydney Memory and Ageing Study.
While prevalence of MCI and different MCI subtypes remains relatively stable across all assessments, reversion from MCI and transitions between different MCI subtypes were common. Up to 46.5% of participants classified with MCI at baseline reverted at some point during follow-up. The majority (83.8%) of participants with incident dementia were diagnosed with MCI 2 years prior to their dementia diagnosis. Both reverters and participants with stable MCI were at an increased risk of progression to dementia compared to those without MCI at baseline (HR 6.4,
= 0.02, and HR 24.7,
< 0.001, respectively); however, the risk of dementia in participants with MCI who did not revert was higher than in reverters (HR 2.5,
= 0.01). This effect was specific to amnestic subtypes (MCI reverters vs nonreverters: amnestic MCI HR 3.3,
= 0.006; nonamnestic MCI: HR 1.3,
= 0.67).
Our findings indicate that the relevance of reversion for progression risk depends on the MCI subtype. Subtype specificity and longitudinal characterization are required for the reliable identification of individuals at high risk of developing dementia.
Abstract
Background
Few studies have compared gait speed and its correlates among different ethnogeographic regions. The goals of this study were to describe usual and rapid gait speed, and identify ...their correlates across Australian, Asian, and African countries.
Methods
We used data from 6 population-based cohorts of adults aged 65+ from 6 countries and 3 continents (N = 6 472), with samples ranging from 231 to 1 913. All cohorts are members of the Cohort Studies of Memory in an International Consortium collaboration. We investigated whether clinical (body mass index BMI, hypertension, stroke, apolipoprotein status), psychological (cognition, mood, general health), and behavioral factors (smoking, drinking, physical activity) correlated with usual (N = 4 cohorts) and rapid gait speed (N = 3 cohorts) similarly across cohorts. Regression models were controlled for age, sex, and education, and were sex-stratified.
Results
Age- and sex-standardized usual gait speed means ranged from 0.61 to 1.06 m/s and rapid gait speed means ranged from 1.16 to 1.64 m/s. Lower BMI and better cognitive function consistently correlated with faster gait speed in all cohorts. Less consistently, not having hypertension and greater physical activity engagement were associated with faster gait speed. Associations with mood, smoking, and drinking were largely nonsignificant. These patterns were not attenuated by demographics. There was limited evidence that the associations differed by sex, except physical activity, where the greater intensity was associated with usual gait among men but not women.
Conclusions
This study is among the first to describe the usual and rapid gait speeds across older adults in Africa, Asia, and Australia.
PDF
