A proof-of-concept study applying machine learning methods to putative risk factors for eating disorders: results from the multi-centre European project on healthy eating.

BACKGROUND: Despite a wide range of proposed risk factors and theoretical models, prediction of eating disorder (ED) onset remains poor. This study undertook the first comparison of two machine learning (ML) approaches [penalised logistic regression (LASSO), and prediction rule ensembles (PREs)] to conventional logistic regression (LR) models to enhance prediction of ED onset and differential ED diagnoses from a range of putative risk factors. METHOD: Data were part of a European Project and comprised 1402 participants, 642 ED patients [52% with anorexia nervosa (AN) and 40% with bulimia nervosa (BN)] and 760 controls. The Cross-Cultural Risk Factor Questionnaire, which assesses retrospectively a range of sociocultural and psychological ED risk factors occurring before the age of 12 years (46 predictors in total), was used. RESULTS: All three statistical approaches had satisfactory model accuracy, with an average area under the curve (AUC) of 86% for predicting ED onset and 70% for predicting AN v. BN. Predictive performance was greatest for the two regression methods (LR and LASSO), although the PRE technique relied on fewer predictors with comparable accuracy. The individual risk factors differed depending on the outcome classification (EDs v. non-EDs and AN v. BN). CONCLUSIONS: Even though the conventional LR performed comparably to the ML approaches in terms of predictive accuracy, the ML methods produced more parsimonious predictive models. ML approaches offer a viable way to modify screening practices for ED risk that balance accuracy against participant burden.

Long-term use of pharmacological treatment in Alzheimer's disease: a retrospective cohort study in real-world clinical practice.

PURPOSE: To assess the impact of long-term use of different drugs commonly prescribed in Alzheimer's disease (AD) on its clinical course and to identify clinical and therapeutic factors associated with a delay in AD progression. METHODS: We retrospectively enrolled 50 patients visited at the Neurology Unit, Careggi University Hospital (Florence), followed for at least 24 months. AD diagnosis was made according to clinical diagnostic criteria for probable/possible AD dementia, always supported at least by one biomarker. Clinical features, MMSE scores evaluated at diagnosis and every 6 months, and AD drugs used for at least 6 months, were recorded. Cox regression analysis was performed to estimate the hazard ratio (HR) for AD progression, assuming as the "final event," the progression to a more severe disease stage, defined as the achievement of an MMSE score less than 10. RESULTS: At baseline, the median MMSE score was 22. During follow-up (median of 41 months), 56% of patients progressed to a more severe disease stage. The use of memantine, either alone (HR 0.24; 95% CI 0.09-0.60) or combined with acetylcholinesterase inhibitors (HR 0.35; 95% CI 0.14-0.88) and a higher MMSE score at baseline (HR 0.82; 95% CI 0.70-0.96) were associated with a significantly lower risk of AD progression. CONCLUSION: Nowadays, effective disease-modifying therapy for AD is missing. Nevertheless, when the diagnosis is established, our results support the advantage of long-term use of available pharmacological treatments, especially in combination, in delaying AD progression to its more severe disease stage.

PER2 C111G polymorphism, cognitive reserve and cognition in subjective cognitive decline and mild cognitive impairment: a 10-year follow-up study.

BACKGROUND AND PURPOSE: CLOCK and PER2 genes have been implicated in sleep-wake cycle alterations and neurodegenerative diseases. Our aim was to evaluate the effect of CLOCK T3111C and PER2 C111G on cognitive functioning in subjective cognitive decline (SCD) patients and mild cognitive impairment (MCI) patients at the baseline of a longitudinal study, and the effect of these two polymorphisms on the progression to Alzheimer's disease (AD) of the two groups. METHODS: Sixty-eight subjects (41 SCD and 27 MCI) who underwent clinical evaluation, neuropsychological assessment, CLOCK and PER2 genotyping at baseline and neuropsychological follow-up every 2 years for a mean time of 10 years were included. Subjects who developed AD (SCD-c and MCI-c) and non-converters (SCD-nc, MCI-nc) were considered. RESULTS: CLOCK T3111C was detected in 47% of cases (21 SCD, 11 MCI) and PER2 C111G in 19% of cases (eight SCD and five MCI). PER2 G carriers presented lower premorbid intelligence score (P = 0.049), fewer years of education (P = 0.007) and a lower frequency of family history of AD (P = 0.04) than G non-carriers. MCI PER2 G carriers had worse performance in tests assessing memory, executive function, language and visuospatial abilities at baseline. During follow-up, two SCD and 15 MCI subjects progressed to AD: both of the SCD-c subjects presented the PER2 G allele, while none of the SCD PER2 G non-carriers converted to AD (P = 0.003). CONCLUSION: PER2 seems to have a role in cognitive reserve and cognition in SCD and MCI patients. Nevertheless, further studies are needed to assess the role of PER2 C111G on the risk of progression to AD.

Assessing the effectiveness of subjective cognitive decline plus criteria in predicting the progression to Alzheimer's disease: an 11-year follow-up study.

BACKGROUND AND PURPOSE: Subjective cognitive decline (SCD) is a self-experienced decline in cognitive capacity with normal performance on standardized cognitive tests and has been shown to increase the risk of Alzheimer's disease (AD). SCD could also be related to other conditions such as normal aging, psychiatric, neurological or medical disorders. The SCD Initiative proposed a set of features (SCD-plus) that increase the likelihood of preclinical AD in individuals with SCD. Our aim was to assess the effect of these features on the risk of conversion from SCD to AD. METHODS: In total 150 SCD subjects who underwent extensive neuropsychological investigation, assessment of cognitive complaints and apolipoprotein E (ApoE) genotyping at baseline and clinical-neuropsychological follow-up for a mean time of 11 years were included. RESULTS: During the follow-up, 20 subjects developed AD. Considering SCD-plus features, age at onset ≥60 years and ApoE ε4 significantly increased the risk of conversion from SCD to AD. When our sample was stratified into three groups (no risk factor, one risk factor, two risk factors), the proportion of conversion was statistically significantly different between the three groups. CONCLUSIONS: Our model allows the risk of AD to be stratified in patients experiencing SCD according to age at onset and ApoE genotype.

Incomplete penetrance in familial Alzheimer's disease with PSEN1 Ala260Gly mutation.

Presenilin1 (PSEN1) gene is the most common known genetic cause of early-onset familial Alzheimer's disease. We describe an Italian family with the known p.Ala260Gly mutation in PSEN1 gene. The presence of an asymptomatic 64-year-old male carrying the mutation provides evidence of a possible incomplete penetrance leading to a wider range of age at onset. In order to evaluate whether or not epigenetic modifications could contribute to the phenotypic heterogeneity, we assessed global DNA methylation levels which resulted significantly higher in the three females than in their presymptomatic brother. The study suggests that DNA methylation can contribute to slowing down or possibly protecting from the manifestation of symptoms even in monogenic diseases, emphasizing the great complexity of familial Alzheimer's disease.

Italian Frontotemporal Dementia Network (FTD Group-SINDEM): sharing clinical and diagnostic procedures in Frontotemporal Dementia in Italy.

In the prospect of improved disease management and future clinical trials in Frontotemporal Dementia, it is desirable to share common diagnostic procedures. To this aim, the Italian FTD Network, under the aegis of the Italian Neurological Society for Dementia, has been established. Currently, 85 Italian Centers involved in dementia care are part of the network. Each Center completed a questionnaire on the local clinical procedures, focused on (1) clinical assessment, (2) use of neuroimaging and genetics; (3) support for patients and caregivers; (4) an opinion about the prevalence of FTD. The analyses of the results documented a comprehensive clinical and instrumental approach to FTD patients and their caregivers in Italy, with about 1,000 newly diagnosed cases per year and 2,500 patients currently followed by the participating Centers. In analogy to other European FTD consortia, future aims will be devoted to collect data on epidemiology of FTD and its subtypes and to provide harmonization of procedures among Centers.

Clinical heterogeneity in Italian patients with amyotrophic lateral sclerosis.

Amyotrophic lateral sclerosis (ALS) is a rare and devastating neurodegenerative disorder. The majority of cases are sporadic ALS (SALS), with 5-10% being familial ALS (FALS), and are inherited mostly as autosomal dominant. Mutations in Cu/Zn superoxide dismutase (SOD1) and the TAR DNA-binding protein (TARDBP) gene are the most commonly known cause of ALS. We analyzed these genes in 61 Italian ALS patients using high-resolution melting analysis to confirm the role of SOD1 and TARDBP in the physiopathology of ALS. The screenings showed a single mutation in SOD1 (Asp109Tyr) and three in TARBDP (Ala382Thr, Gly295Ser, Gly294Val) in five unrelated ALS patients. This report enlarges the spectrum of clinical phenotypes associated with genetic mutations in SOD1 and TARDBP genes confirming the variability of phenotypes associated with the same mutation and emphasizes the importance of genetic analysis. The different genotype-phenotype correlations suggest the implication of other factors possibly influencing clinical manifestation of the disease, such as an epigenetic or epistatic effect with other genes not yet identified.

APOE and Alzheimer disease: a major gene with semi-dominant inheritance.

Apolipoprotein E (APOE) dependent lifetime risks (LTRs) for Alzheimer Disease (AD) are currently not accurately known and odds ratios alone are insufficient to assess these risks. We calculated AD LTR in 7351 cases and 10 132 controls from Caucasian ancestry using Rochester (USA) incidence data. At the age of 85 the LTR of AD without reference to APOE genotype was 11% in males and 14% in females. At the same age, this risk ranged from 51% for APOE44 male carriers to 60% for APOE44 female carriers, and from 23% for APOE34 male carriers to 30% for APOE34 female carriers, consistent with semi-dominant inheritance of a moderately penetrant gene. Using PAQUID (France) incidence data, estimates were globally similar except that at age 85 the LTRs reached 68 and 35% for APOE 44 and APOE 34 female carriers, respectively. These risks are more similar to those of major genes in Mendelian diseases, such as BRCA1 in breast cancer, than those of low-risk common alleles identified by recent GWAS in complex diseases. In addition, stratification of our data by age groups clearly demonstrates that APOE4 is a risk factor not only for late-onset but for early-onset AD as well. Together, these results urge a reappraisal of the impact of APOE in Alzheimer disease.

EFNS-ENS Guidelines on the diagnosis and management of disorders associated with dementia.

BACKGROUND AND OBJECTIVES: The last version of the EFNS dementia guidelines is from 2007. In 2010, the revised guidelines for Alzheimer's disease (AD) were published. The current guidelines involve the revision of the dementia syndromes outside of AD, notably vascular cognitive impairment, frontotemporal lobar degeneration, dementia with Lewy bodies, corticobasal syndrome, progressive supranuclear palsy, Parkinson's disease dementia, Huntington's disease, prion diseases, normal-pressure hydrocephalus, limbic encephalitis and other toxic and metabolic disorders. The aim is to present a peer-reviewed evidence-based statement for the guidance of practice for clinical neurologists, geriatricians, psychiatrists and other specialist physicians responsible for the care of patients with dementing disorders. It represents a statement of minimum desirable standards for practice guidance. METHODS: The task force working group reviewed evidence from original research articles, meta-analyses and systematic reviews, published by June 2011. The evidence was classified (I, II, III, IV) and consensus recommendations graded (A, B, or C) according to the EFNS guidance. Where there was a lack of evidence, but clear consensus, good practice points were provided. RESULTS AND CONCLUSIONS: New recommendations and good practice points are made for clinical diagnosis, blood tests, neuropsychology, neuroimaging, electroencephalography, cerebrospinal fluid (CSF) analysis, genetic testing, disclosure of diagnosis, treatment of behavioural and psychological symptoms in dementia, legal issues, counselling and support for caregivers. All recommendations were revised as compared with the previous EFNS guidelines. The specialist neurologist together with primary care physicians play an important role in the assessment, interpretation and treatment of symptoms, disability and needs of dementia patients.

ApolipoproteinE epsilon 4 allele is not associated with disease course and severity in multiple sclerosis.

BACKGROUND: No study has assessed the association between apolipoproteinE (APOE) and multiple sclerosis (MS) forms grouped by also taking into account cognitive performance. AIMS OF THE STUDY: To assess the relationship between APOE and disease course, particularly focusing on benign MS (BMS), defined as also including cognitive preservation. METHODS: In 173 consecutive patients, we assessed the association between APOE and MS course and severity. RESULTS: Twenty-nine APOE-epsilon4 carriers were identified. The epsilon4 allele was not associated with BMS. Moreover, it was associated neither with other disease courses nor with the time to reach disability milestones and secondary progression. CONCLUSION: Although plausible, the association between APOE and MS course (particularly with BMS defined by including cognitive preservation) and disease severity remains controversial.

Codon 129 polymorphism of prion protein gene in sporadic Alzheimer's disease.

Codon 129 polymorphism of the prion protein gene represents a major genetic risk factor for Creutzfeldt-Jakob disease (CJD). Both CJD and Alzheimer's disease (AD) are brain amyloidoses and it would be possible that codon 129 polymorphism plays a role in the susceptibility to AD. In order to investigate this polymorphism in AD the distribution of polymorphic codon 129 of the PRNP gene in 194 probable AD and 124 controls selected in Italy and 109 neuropathologically verified AD and 58 matched controls recruited in the USA was studied. No significant association was found for the PRNP polymorphism in AD compared to controls either in Probable or in Definite AD series even after stratification for APOE polymorphism. This study does not support a role of PRNP polymorphism as a susceptibility factor for AD.

The different apoptotic potential of the p53 codon 72 alleles increases with age and modulates in vivo ischaemia-induced cell death.

A common arginine to proline polymorphism is harboured at codon 72 of the human p53 gene. In this investigation, we found that fibroblasts and lymphocytes isolated from arginine allele homozygote centenarians and sexagenarians (Arg+) undergo an oxidative-stress-induced apoptosis at a higher extent than cells obtained from proline allele carriers (Pro+). At variance, the difference in apoptosis susceptibility between Arg+ and Pro+ is not significant when cells from 30-year-old people are studied. Further, we found that Arg+ and Pro+ cells from centenarians differ in the constitutive levels of p53 protein and p53/MDM2 complex, as well as in the levels of oxidative stress-induced p53/Bcl-xL complex and mitochondria-localised p53. Consistently, all these differences are less evident in cells from 30-year-old people. Finally, we investigated the in vivo functional relevance of the p53 codon 72 genotype in a group of old patients (66-99 years of age) affected by acute myocardial ischaemia, a clinical condition in which in vivo cell death occurs. We found that Arg+ patients show increased levels of Troponin I and CK-MB, two serum markers that correlate with the extent of the ischaemic damage in comparison to Pro+ patients. In conclusion, these data suggest that p53 codon 72 polymorphism contributes to a genetically determined variability in apoptotic susceptibility among old people, which has a potentially relevant role in the context of an age-related pathologic condition, such as myocardial ischaemia.

Association between 5-HT(2A) receptor polymorphism and psychotic symptoms in Alzheimer's disease.

BACKGROUND: The aim of this study is to analyze the segregation of the 102T/C polymorphism in the serotonin 2A receptor gene in patients affected by sporadic and familial Alzheimer's disease (FAD) with and without psychotic symptoms. METHODS: The polymorphism was analyzed in 275 subjects. A semistructured interview was used to obtain information about delusions, hallucinations, and other specific behavioral signs occurring during the disease. RESULTS: Fifty-two percent of AD patients with psychotic symptoms were homozygous for the C102 allele, as compared with 6.9% of AD patients without psychosis. Similarly, the C102/C102 genotype was significantly more frequent in FAD patients with psychosis than in FAD patients without (46.5% vs. 7.8%). CONCLUSIONS: Our data strongly confirm and extend to FAD previous studies suggesting that the genetic variation at this locus is associated with prominent psychotic features in AD and that the 102C allele could play an important role in late-onset AD.

Genetic and clinical analysis of spinocerebellar ataxia type 8 repeat expansion in Italy.

BACKGROUND: The spinocerebellar ataxias (SCAs) are clinically heterogeneous disorders caused by triplet repeat expansions in the sequence of specific disease genes. Spinocerebellar ataxia type 8 (SCA8), originally described in a family characterized by pure cerebellar ataxia with slow disease progression, presents with expansion of combined CTA/CTG repeats. OBJECTIVE: To perform SCA8 repeat expansion analysis in a heterogeneous group of ataxic patients, to determine the prevalence of this mutation in our patients and establish the frequency of expanded CTA/CTG repeats in a large group of control subjects. PATIENTS: One hundred sixty-seven patients affected by sporadic, autosomal dominant and recessive hereditary ataxia were clinically examined and analyzed for SCA8 expansion. We further studied 161 control subjects and 125 patients with psychiatric disorders. RESULTS: We found abnormally expanded CTA/CTG repeats in 5 ataxic patients, 3 of them characterized by pure cerebellar ataxia. One patient had vitamin E deficiency and 1 patient with a sporadic case was affected by gluten ataxia. No evidence of expanded alleles was found in healthy control subjects and in patients with psychiatric disorders. CONCLUSIONS: Our data support the evidence that CTG expansions may be linked to SCA8, since the pathogenic expansions have been found only among patients with genetically unidentified forms of hereditary and sporadic ataxia. Patients carrying expanded alleles present peculiar phenotypic features, thus suggesting that unknown additional factors could probably predispose to the disease.

Genetic risk factors in familial Alzheimer's disease.

In the last 10 years significant progress has been made to describe and identify the underlying biological mechanisms that cause the different manifestation of Alzheimer's disease. Since the first report of a possible locus on chromosome 21 in a small group of families with early onset familial Alzheimer's disease (FAD), considerable progress has been made. Results from linkage analysis and gene sequencing has provided evidence that a minority of early onset FAD families develops the disease as a result of mutations in the gene coding for the Abeta-amyloid precursor protein, and that mutations in presenilin 1 and 2 genes account for a larger subgroup of early onset families. Several other early onset FAD families are clearly not linked to any of these loci, suggesting that other genetic risk factors may exist. Recent genome-wide scanning studies have revealed the existence of a new locus on chromosome 12, which, together with inheritance of the epsilon4 allele of apolipoprotein E gene, on chromosome 19, represent the most important genetic factors associated with an increased risk of developing the disease in late onset FAD families.

Molecular genetics of Alzheimer's disease in Italian families.

We screened 11 families from different regions of Italy by direct sequencing of exon 17 of the APP gene. Two unrelated families carried the APP717 mutation segregating with the disease. These two families originate from two Italian regions which are considered genetically separate. Published studies have demonstrated the presence of the APP717 Val-->Ile mutation in kindreds of British or Japanese origin with early onset familial Alzheimer's disease. These data suggest that the APP717 mutation is not confined to islander families which may share common founders. From the molecular genetic point of view we also did linkage analysis. Several families, in fact, have not shown a linkage with chromosome 21 and the resolution of this dilemma required investigation of those pedigrees both with additional markers from chromosome 21 and with markers from other chromosomes.

Alzheimer's disease and apolipoprotein E in Italy.

Recent studies have provided evidence of association of apolipoprotein E (ApoE) epsilon 4 allele and late onset familial and sporadic Alzheimer's disease (AD). Epidemiological studies have established allelic variation at the ApoE locus. We have analyzed the ApoE gene polymorphism in a sample of 416 Italian subjects. Our data confirm a significant association between epsilon 4 allele and sporadic AD. The frequency of epsilon 4 allele in early onset familial AD patients was comparable to control values suggesting that epsilon 4 allele does not represent a risk factor for early onset familial AD (EOFAD). Moreover, we found a not-previously reported association between ApoE epsilon 2 allele and sporadic AD and EOFAD. We included in this study two EOFAD families with the APP717 Val-->Ile mutation in the Amyloid Precursor Protein (APP) gene on chromosome 21. In any of the EOFAD families there was a significant effect of the ApoE genotype on the age of onset with the exception of one of the two mutated EOFAD families in which the 2 allele delays the age of onset.

ApoE allele frequencies in Italian sporadic and familial Alzheimer's disease.

Recent studies have provided evidence of association of apolipoprotein E (ApoE) epsilon 4 allele and late onset familial and sporadic Alzheimer's disease (AD). Epidemiological studies have established allelic variation at the ApoE locus. We have analyzed the ApoE gene polymorphism in a sample of 446 Italian subjects. Our data confirm a significant association between epsilon 4 allele and sporadic AD. The frequency of epsilon 4 allele in early onset familial AD patients was comparable to control values suggesting that epsilon 4 allele does not represent a risk factor for early onset familial AD (EOFAD). Moreover, we found a not previously reported association between ApoE epsilon 2 allele and sporadic AD and EOFAD.