NLS: An accurate and yet easy-to-interpret prediction method.

Over the last years, the predictive power of supervised machine learning (ML) has undergone impressive advances, achieving the status of state of the art and super-human level in some applications. However, the employment rate of ML models in real-life applications is much slower than one would expect. One of the downsides of using ML solution-based technologies is the lack of user trust in the produced model, which is related to the black-box nature of these models. To leverage the application of ML models, the generated predictions should be easy to interpret while maintaining a high accuracy. In this context, we develop the Neural Local Smoother (NLS), a neural network architecture that yields accurate predictions with easy-to-obtain explanations. The key idea of NLS is to add a smooth local linear layer to a standard network. We show experiments that indicate that NLS leads to a predictive power that is comparable to state-of-the-art machine learning models, but that at the same time is easier to interpret.

Selecting Genetic Variants and Interactions Associated with Amyotrophic Lateral Sclerosis: A Group LASSO Approach.

Amyotrophic lateral sclerosis (ALS) is a multi-system neurodegenerative disease that affects both upper and lower motor neurons, resulting from a combination of genetic, environmental, and lifestyle factors. Usually, the association between single-nucleotide polymorphisms (SNPs) and this disease is tested individually, which leads to the testing of multiple hypotheses. In addition, this classical approach does not support the detection of interaction-dependent SNPs. We applied a two-step procedure to select SNPs and pairwise interactions associated with ALS. SNP data from 276 ALS patients and 268 controls were analyzed by a two-step group LASSO in 2000 iterations. In the first step, we fitted a group LASSO model to a bootstrap sample and a random subset of predictors (25%) from the original data set aiming to screen for important SNPs and, in the second step, we fitted a hierarchical group LASSO model to evaluate pairwise interactions. An in silico analysis was performed on a set of variables, which were prioritized according to their bootstrap selection frequency. We identified seven SNPs (rs16984239, rs10459680, rs1436918, rs1037666, rs4552942, rs10773543, and rs2241493) and two pairwise interactions (rs16984239:rs2118657 and rs16984239:rs3172469) potentially involved in nervous system conservation and function. These results may contribute to the understanding of ALS pathogenesis, its diagnosis, and therapeutic strategy improvement.

The Latent Dirichlet Allocation model with covariates (LDAcov): A case study on the effect of fire on species composition in Amazonian forests.

Understanding and predicting the effect of global change phenomena on biodiversity is challenging given that biodiversity data are highly multivariate, containing information from tens to hundreds of species in any given location and time. The Latent Dirichlet Allocation (LDA) model has been recently proposed to decompose biodiversity data into latent communities. While LDA is a very useful exploratory tool and overcomes several limitations of earlier methods, it has limited inferential and predictive skill given that covariates cannot be included in the model. We introduce a modified LDA model (called LDAcov) which allows the incorporation of covariates, enabling inference on the drivers of change of latent communities, spatial interpolation of results, and prediction based on future environmental change scenarios. We show with simulated data that our approach to fitting LDAcov is able to estimate well the number of groups and all model parameters. We illustrate LDAcov using data from two experimental studies on the long-term effects of fire on southeastern Amazonian forests in Brazil. Our results reveal that repeated fires can have a strong impact on plant assemblages, particularly if fuel is allowed to build up between consecutive fires. The effect of fire is exacerbated as distance to the edge of the forest decreases, with small-sized species and species with thin bark being impacted the most. These results highlight the compounding impacts of multiple fire events and fragmentation, a scenario commonly found across the southern edge of Amazon. We believe that LDAcov will be of wide interest to scientists studying the effect of global change phenomena on biodiversity using high-dimensional datasets. Thus, we developed the R package LDAcov to enable the straightforward use of this model.

[Machine learning for predictive analyses in health: an example of an application to predict death in the elderly in São Paulo, Brazil]. / Machine learning para análises preditivas em saúde: exemplo de aplicação para predizer óbito em idosos de São Paulo, Brasil.

This study aims to present the stages related to the use of machine learning algorithms for predictive analyses in health. An application was performed in a database of elderly residents in the city of São Paulo, Brazil, who participated in the Health, Well-Being, and Aging Study (SABE) (n = 2,808). The outcome variable was the occurrence of death within five years of the elder's entry into the study (n = 423), and the predictors were 37 variables related to the elder's demographic, socioeconomic, and health profile. The application was organized according to the following stages: division of data in training (70%) and testing (30%), pre-processing of the predictors, learning, and assessment of the models. The learning stage used 5 algorithms to adjust the models: logistic regression with and without penalization, neural networks, gradient boosted trees, and random forest. The algorithms' hyperparameters were optimized by 10-fold cross-validation to select those corresponding to the best models. For each algorithm, the best model was assessed in test data via area under the ROC curve (AUC) and related measures. All the models presented AUC ROC greater than 0.70. For the three models with the highest AUC ROC (neural networks and logistic regression with LASSO penalization and without penalization, respectively), quality measures of the predicted probability were also assessed. The expectation is that with the increased availability of data and trained human capital, it will be possible to develop predictive machine learning models with the potential to help health professionals make the best decisions.

Este estudo objetiva apresentar as etapas relacionadas à utilização de algoritmos de machine learning para análises preditivas em saúde. Para isso, foi realizada uma aplicação com base em dados de idosos residentes no Município de São Paulo, Brasil, participantes do estudo Saúde Bem-estar e Envelhecimento (SABE) (n = 2.808). A variável resposta foi representada pela ocorrência de óbito em até cinco anos após o ingresso do idoso no estudo (n = 423), e os preditores, por 37 variáveis relacionadas ao perfil demográfico, socioeconômico e de saúde do idoso. A aplicação foi organizada de acordo com as seguintes etapas: divisão dos dados em treinamento (70%) e teste (30%), pré-processamento dos preditores, aprendizado e avaliação de modelos. Na etapa de aprendizado, foram utilizados cinco algoritmos para o ajuste de modelos: regressão logística com e sem penalização, redes neurais, gradient boosted trees e random forest. Os hiperparâmetros dos algoritmos foram otimizados por validação cruzada 10-fold, para selecionar aqueles correspondentes aos melhores modelos. Para cada algoritmo, o melhor modelo foi avaliado em dados de teste por meio da área abaixo da curva (AUC) ROC e medidas relacionadas. Todos os modelos apresentaram AUC ROC superior a 0,70. Para os três modelos com maior AUC ROC (redes neurais e regressão logística com penalização de lasso e sem penalização, respectivamente), foram também avaliadas medidas de qualidade da probabilidade predita. Espera-se que, com o aumento da disponibilidade de dados e de capital humano capacitado, seja possível desenvolver modelos preditivos de machine learning com potencial para auxiliar profissionais de saúde na tomada de melhores decisões.

El objetivo de este estudio fue presentar las etapas relacionadas con la utilización de algoritmos de machine learning para análisis predictivos en salud. Para tal fin, se realizó una aplicación en base a datos de ancianos residentes en el Municipio de São Paulo, Brasil, participantes en el estudio Salud Bienestar y Envejecimiento (SABE) (n = 2.808). La variable respuesta se representó mediante la ocurrencia de óbito en hasta 5 años tras la inclusión del anciano en el estudio (n = 423), y los predictores fueron representados por 37 variables relacionadas con el perfil demográfico, socioeconómico y de salud del anciano. El aplicación se organizó según las siguientes etapas: división de los datos en formación (70%) y test (30%), pre-procesamiento de los predictores, aprendizaje y evaluación de modelos. En la etapa de aprendizaje, se utilizaron cinco algoritmos para el ajuste de modelos: regresión logística con y sin penalización, redes neuronales, gradient boosted trees y random forest. Los hiperparámetros de los algoritmos se optimizaron mediante una validación cruzada 10-fold, para seleccionar aquellos correspondientes a los mejores modelos. Para cada algoritmo, el mejor modelo se evaluó con datos de la prueba del área debajo de la curva (AUC) ROC y medidas relacionadas. Todos los modelos presentaron AUC ROC superior a 0,70. Para los tres modelos con mayor AUC ROC (redes neuronales y regresión logística con penalización de Lasso y sin penalización, respectivamente) también se evaluaron medidas de calidad de la probabilidad pronosticada. Se espera que, con el aumento de la disponibilidad de datos y de capital humano capacitado, sea posible desarrollar modelos predictivos de machine learning con potencial para ayudar a profesionales de salud en la toma de mejores decisiones.

Avian haemosporidians in the cattle egret (Bubulcus ibis) from central-western and southern Africa: High diversity and prevalence.

We described the geographic distribution of 82 haemosporidian lineages (Plasmodium, Haemoproteus, and Leucocytozoon) in the cattle egret sampled in five countries in central-western and southern Africa. Seventy-three lineages have not previously been reported. We determined the prevalence of three haemosporidians in the samples. We investigated the influence of the internal environment of the host and environmental variables on the Plasmodium diversity and whether environmental variables may explain spatial variations in the prevalence of Plasmodium. We screened DNA from 509 blood samples from nestlings in 15 African colonies for infection by sequencing the cytochrome b gene of parasites. The molecular phylogenetic analysis was performed using Bayesian methods and including sequences from the MalAvi and GeneBank databases. We found 62 new Plasmodium lineages in a clade with MYCAME02, which is a lineage described in waterbirds and recently identified in birds of prey as Plasmodium paranucleophilum. Two Haemoproteus lineages identified in cattle egret formed a distinct group with Haemoproteus catharti and MYCAMH1 (Haemoproteus spp.). Seven Leucocytozoon lineages found in the cattle egret clustered with Leucocytozoon californicus. We found different Plasmodium diversities among the colonies sampled, demonstrating that the internal environment of the host is not the primary determinant of diversity. A linear mixed-effects multivariate model showed that precipitation was positively associated with Plasmodium diversity when controlling for the effects of temperature, colony composition (mixed and non-mixed species) and country. Moreover, a generalized mixed model showed that temperature was positively associated with the prevalence of Plasmodium when controlling for precipitation, elevation and country. We conclude that the cattle egret is a good model for future haemosporidian studies, as we found a significant number of new lineages in this host, which occupies regions with different climate characteristics where environmental variables exert an influence on the diversity and prevalence of Plasmodium.

Machine learning para análises preditivas em saúde: exemplo de aplicação para predizer óbito em idosos de São Paulo, Brasil / Machine learning for predictive analyses in health: an example of an application to predict death in the elderly in São Paulo, Brazil / Machine learning para análisis predictivos en salud: ejemplo de aplicación para la predicción de óbitos en ancianos de São Paulo, Brasil

Este estudo objetiva apresentar as etapas relacionadas à utilização de algoritmos de machine learning para análises preditivas em saúde. Para isso, foi realizada uma aplicação com base em dados de idosos residentes no Município de São Paulo, Brasil, participantes do estudo Saúde Bem-estar e Envelhecimento (SABE) (n = 2.808). A variável resposta foi representada pela ocorrência de óbito em até cinco anos após o ingresso do idoso no estudo (n = 423), e os preditores, por 37 variáveis relacionadas ao perfil demográfico, socioeconômico e de saúde do idoso. A aplicação foi organizada de acordo com as seguintes etapas: divisão dos dados em treinamento (70%) e teste (30%), pré-processamento dos preditores, aprendizado e avaliação de modelos. Na etapa de aprendizado, foram utilizados cinco algoritmos para o ajuste de modelos: regressão logística com e sem penalização, redes neurais, gradient boosted trees e random forest. Os hiperparâmetros dos algoritmos foram otimizados por validação cruzada 10-fold, para selecionar aqueles correspondentes aos melhores modelos. Para cada algoritmo, o melhor modelo foi avaliado em dados de teste por meio da área abaixo da curva (AUC) ROC e medidas relacionadas. Todos os modelos apresentaram AUC ROC superior a 0,70. Para os três modelos com maior AUC ROC (redes neurais e regressão logística com penalização de lasso e sem penalização, respectivamente), foram também avaliadas medidas de qualidade da probabilidade predita. Espera-se que, com o aumento da disponibilidade de dados e de capital humano capacitado, seja possível desenvolver modelos preditivos de machine learning com potencial para auxiliar profissionais de saúde na tomada de melhores decisões.

This study aims to present the stages related to the use of machine learning algorithms for predictive analyses in health. An application was performed in a database of elderly residents in the city of São Paulo, Brazil, who participated in the Health, Well-Being, and Aging Study (SABE) (n = 2,808). The outcome variable was the occurrence of death within five years of the elder's entry into the study (n = 423), and the predictors were 37 variables related to the elder's demographic, socioeconomic, and health profile. The application was organized according to the following stages: division of data in training (70%) and testing (30%), pre-processing of the predictors, learning, and assessment of the models. The learning stage used 5 algorithms to adjust the models: logistic regression with and without penalization, neural networks, gradient boosted trees, and random forest. The algorithms' hyperparameters were optimized by 10-fold cross-validation to select those corresponding to the best models. For each algorithm, the best model was assessed in test data via area under the ROC curve (AUC) and related measures. All the models presented AUC ROC greater than 0.70. For the three models with the highest AUC ROC (neural networks and logistic regression with LASSO penalization and without penalization, respectively), quality measures of the predicted probability were also assessed. The expectation is that with the increased availability of data and trained human capital, it will be possible to develop predictive machine learning models with the potential to help health professionals make the best decisions.

El objetivo de este estudio fue presentar las etapas relacionadas con la utilización de algoritmos de machine learning para análisis predictivos en salud. Para tal fin, se realizó una aplicación en base a datos de ancianos residentes en el Municipio de São Paulo, Brasil, participantes en el estudio Salud Bienestar y Envejecimiento (SABE) (n = 2.808). La variable respuesta se representó mediante la ocurrencia de óbito en hasta 5 años tras la inclusión del anciano en el estudio (n = 423), y los predictores fueron representados por 37 variables relacionadas con el perfil demográfico, socioeconómico y de salud del anciano. El aplicación se organizó según las siguientes etapas: división de los datos en formación (70%) y test (30%), pre-procesamiento de los predictores, aprendizaje y evaluación de modelos. En la etapa de aprendizaje, se utilizaron cinco algoritmos para el ajuste de modelos: regresión logística con y sin penalización, redes neuronales, gradient boosted trees y random forest. Los hiperparámetros de los algoritmos se optimizaron mediante una validación cruzada 10-fold, para seleccionar aquellos correspondientes a los mejores modelos. Para cada algoritmo, el mejor modelo se evaluó con datos de la prueba del área debajo de la curva (AUC) ROC y medidas relacionadas. Todos los modelos presentaron AUC ROC superior a 0,70. Para los tres modelos con mayor AUC ROC (redes neuronales y regresión logística con penalización de Lasso y sin penalización, respectivamente) también se evaluaron medidas de calidad de la probabilidad pronosticada. Se espera que, con el aumento de la disponibilidad de datos y de capital humano capacitado, sea posible desarrollar modelos predictivos de machine learning con potencial para ayudar a profesionales de salud en la toma de mejores decisiones.

Considerations for monitoring population trends of colonial waterbirds using the effective number of breeders and census estimates.

Detecting trends in population size fluctuations is a major focus in ecology, evolution, and conservation biology. Populations of colonial waterbirds have been monitored using demographic approaches to determine annual census size (Na). We propose the addition of genetic estimates of the effective number of breeders (Nb) as indirect measures of the risk of loss of genetic diversity to improve the evaluation of demographics and increase the accuracy of trend estimates in breeding colonies. Here, we investigated which methods of the estimation of Nb are more precise under conditions of moderate genetic diversity, limited sample sizes and few microsatellite loci, as often occurs with natural populations. We used the wood stork as a model species and we offered a workflow that researchers can follow for monitoring bird breeding colonies. Our approach started with simulations using five estimators of Nb and the theoretical results were validated with empirical data collected from breeding colonies settled in the Brazilian Pantanal wetland. In parallel, we estimated census size using a corrected method based on counting active nests. Both in simulations and in natural populations, the approximate Bayesian computation (ABC) and sibship assignment (SA) methods yielded more precise estimates than the linkage disequilibrium, heterozygosity excess, and molecular coancestry methods. In particular, the ABC method performed best with few loci and small sample sizes, while the other estimators required larger sample sizes and at least 13 loci to not underestimate Nb. Moreover, according to our Nb/Na estimates (values were often ≤0.1), the wood stork colonies evaluated could be facing the loss of genetic diversity. We demonstrate that the combination of genetic and census estimates is a useful approach for monitoring natural breeding bird populations. This methodology has been recommended for populations of rare species or with a known history of population decline to support conservation efforts.

Hypertension and frailty in older adults.

The association between hypertension and frailty syndrome in older adults remains unclear. There is scarce information about the prevalence of hypertension among frail elderly patients or on its relationship with frailty. Up to one quarter of frail elderly patients present without comorbidity or disability, yet frailty is a leading cause of death. The knowledge and better control of frailty risk factors could influence prognosis. The present study evaluated: (1) the prevalence of hypertension in robust, prefrail, and frail elderly; and (2) factors that might be associated with frailty including hypertension. A cross-sectional study was conducted in 619 older adults at a university-based outpatient center. Study protocol included sociodemographic data, measures of blood pressure and body mass index, frailty screening according to the internationally validated FRAIL (fatigue, resistance, ambulation, illnesses, and loss of weight) scale, number of comorbidities, drug use assessment, physical activity, cognitive status, and activities of daily living. Ordinal logistic regression was used to evaluate factors associated with frailty. Prevalence of hypertension and frailty was 67.3% and 14.8%, respectively, in the total sample. Hypertension was more prevalent in the prefrail (72.5%) and frail (83%) groups than among controls (51.7%). Hypertension, physical activity, number of prescribed drugs, and cognitive performance were significantly associated with frailty status. Hypertension presented an odds ratio of 1.77 towards frailty (95% confidence interval, 1.21-2.60; P = .002). Hypertension was more prevalent in frail elderly patients and was significantly associated with frailty. Intensive control of hypertension could influence the trajectory of frailty, and this hypothesis should be explored in future prospective clinical trials.

Frailty in older adults with amnestic mild cognitive impairment as a result of Alzheimer's disease: A comparison of two models of frailty characterization.

AIM: To verify the prevalence and presence of frailty markers, and their relationship to cognitive function among older adults with amnestic mild cognitive impairment (aMCI). METHODS: This was an observational study with transversal analyses. Older adults with aMCI as a result of Alzheimer's disease (n = 40) were compared with healthy controls (n = 26) at the Psychogeriatric Outpatient Unit, Institute and Department of Psychiatry, Faculty of Medicine of the University of São Paulo. All participants were submitted to a broad clinical and neuropsychological evaluation. Frailty was evaluated according to the Cardiovascular Health Study (CHS) phenotype and the Edmonton Frail Scale (EFS). MCI was diagnosed by a multidisciplinary consensus according to the Petersen criteria and cerebrospinal fluid analysis for Alzheimer's disease biomarkers. RESULTS: The prevalence of frailty was significantly higher in the aMCI compared with the control group when it was assessed with the EFS (P = 0.047), but not with the CHS (P = 0.255). The prevalence of frailty varied on the criteria used (EFS 7.5%; CHS 30%). The fatigue variable in the CHS (P = 0.036), and the mood (P = 0.019) and functional independence (P = 0.042) variables from the EFS were significantly different between the groups. Visuospatial function (OR 2.405, P = 0.042) was associated with the CHS criteria. CONCLUSION: The identification of frailty features in aMCI appears to depend on the protocol used for evaluation. Visuospatial function showed a higher risk for frailty with the CHS. Geriatr Gerontol Int 2017; 17: 2096-2102.

Screening for Frailty With the FRAIL Scale: A Comparison With the Phenotype Criteria.

BACKGROUND: Reliable and valid frailty screening instruments are lacking. The aim of the present study was to compare the diagnostic properties of the FRAIL-BR with Fried's frailty phenotype (CHS), which has not been done. METHODS: Cross-sectional observational study of 124 older adults aged 60 or older from 2 university-based geriatric outpatient units in the state of São Paulo, Brazil. In ROC analyses, we evaluated different cutoff points and AUC areas of the FRAIL-BR compared with the CHS criteria. Also, components of both diagnostic strategies had head-to-head comparisons whenever possible. RESULTS: The sample was composed mostly of overweight (mean BMI = 29.5 kg/m2) women (83%) with mean age of 78.6 (±7.1) years. Prevalence of frailty varied according to the FRAIL-BR (23.3%) and the CHS criteria (14.5%) (P = .04). A cutoff of 3 points in the FRAIL-BR presented a sensitivity of 28% and specificity of 90% (P = .049). A cutoff of 2 points resulted in a sensitivity of 54% and specificity of 73% (P = .01). Comparisons of 4 FRAIL-BR items (ie, weight loss, aerobic capacity, fatigue, and physical resistance) to the respective CHS components showed an independent diagnostic property of all measures, with the exception for weight loss. CONCLUSION: The FRAIL scale can be used as a screening instrument for frailty (time and cost-effective).

A subtest analysis of The Montreal Cognitive Assessment (MoCA): which subtests can best discriminate between healthy controls, mild cognitive impairment and Alzheimer's disease?

The authors would like to apologise for a typographical error in the discussion of the above mentioned article. In the discussion on page 830 of the article, paragraph 'In the present sample, when we tested the accuracy of the MoCA to discriminate between MCI and healthy participants using ROC curves, the best cut-off score was 24 points, with good sensitivity and specificity ( 92% and 82%, respectively).' Should read: In the present sample, when we tested the accuracy of the MoCA to discriminate between MCI and healthy participants using ROC curves, the best cut-off score was 24 points, with good sensitivity and specificity (83% and 89%, respectively).

A subtest analysis of the Montreal cognitive assessment (MoCA): which subtests can best discriminate between healthy controls, mild cognitive impairment and Alzheimer's disease?

BACKGROUND: It is necessary to continue to explore the psychometric characteristics of key cognitive screening tests such as the Montreal Cognitive Assessment (MoCA) to diagnose cognitive decline as early as possible and to attend to the growing need of clinical trials involving mild cognitive impairment (MCI) participants. The main aim of this study was to assess which MoCA subtests could best discriminate between healthy controls (HC), participants with MCI, and Alzheimer's disease (AD). METHODS: Cross-sectional analysis of 136 elderly with more than four years of education. All participants were submitted to detailed clinical, laboratory, and neuroimaging evaluation. The MoCA, Mini-Mental State Examination (MMSE), the Cambridge Cognitive Examination (CAMCOG), Geriatric Depression Scale (GDS), and Functional Activities Questionnaire (FAQ) were applied to all participants. The MoCA test was not used in the diagnostic procedure. RESULTS: Median MoCA total scores were 27, 23 and 18 for HC, MCI, and AD, respectively (p < 0.001). Word repetition, inverse digits, serial 7, phrases, verbal fluency, abstraction, and word recall discriminated between MCI and HC participants (p < 0.001). The clock drawing, the rhino naming, delayed recall of five words and orientation discriminated between patients with MCI and AD (p < 0.001). A reduced version of the MoCA with only these items did not improve accuracy between MCI and HC (p = 0.076) or MCI and AD (p = 0.119). CONCLUSIONS: Not all MoCA subtests might be fundamental to clinical diagnosis of MCI. The reduced versions of MoCA did not add diagnostic accuracy.

Religion as a protective factor against drug use among Brazilian university students: a national survey.

OBJECTIVE: To investigate the relationship between religiosity and drug use among Brazilian university students. METHODS: This manuscript is part of the "First Nationwide Survey on the Use of Alcohol, Tobacco and Other Drugs among College Students in the 27 Brazilian State Capitals". In this study, 12,595 university students were divided into two groups according to their attendance at religious services: frequent attenders (FR; 39.1%) and non-frequent attenders (NFR; 60.8%). Subsequently, we analyzed their responses to a structured, anonymous questionnaire on drug use and other behaviors. Individual multivariate logistic regression models tested the association between religiosity and drug use (alcohol, tobacco, marijuana and at least one illicit drug). RESULTS: Drug use over the last 30 days was higher among NFR students even after controlling for demographic variables. NFR students were more likely to use alcohol OR = 2.52; 95% CI: 2.08-3.06, tobacco (2.83; 2.09-3.83), marijuana (2.09; 1.39-3.11) and at least one illicit drug (1.42; 1.12-1.79) compared to FR students. CONCLUSION: Religiosity was found to be a strongly protective factor against drug use among Brazilian university students. However, more studies are needed to identify the mechanisms by which religiosity exerts this protective influence.

Religion as a protective factor against drug use among brazilian university students: a national survey

OBJECTIVE: To investigate the relationship between religiosity and drug use among Brazilian university students. METHODS: This manuscript is part of the "First Nationwide Survey on the Use of Alcohol, Tobacco and Other Drugs among College Students in the 27 Brazilian State Capitals". In this study, 12,595 university students were divided into two groups according to their attendance at religious services: frequent attenders (FR; 39.1%) and non-frequent attenders (NFR; 60.8%). Subsequently, we analyzed their responses to a structured, anonymous questionnaire on drug use and other behaviors. Individual multivariate logistic regression models tested the association between religiosity and drug use (alcohol, tobacco, marijuana and at least one illicit drug). RESULTS: Drug use over the last 30 days was higher among NFR students even after controlling for demographic variables. NFR students were more likely to use alcohol OR = 2.52; 95% CI: 2.08-3.06, tobacco (2.83; 2.09-3.83), marijuana (2.09; 1.39-3.11) and at least one illicit drug (1.42; 1.12-1.79) compared to FR students. CONCLUSION: Religiosity was found to be a strongly protective factor against drug use among Brazilian university students. However, more studies are needed to identify the mechanisms by which religiosity exerts this protective influence.

Testing allele homogeneity: the problem of nested hypotheses.

BACKGROUND: The evaluation of associations between genotypes and diseases in a case-control framework plays an important role in genetic epidemiology. This paper focuses on the evaluation of the homogeneity of both genotypic and allelic frequencies. The traditional test that is used to check allelic homogeneity is known to be valid only under Hardy-Weinberg equilibrium, a property that may not hold in practice. RESULTS: We first describe the flaws of the traditional (chi-squared) tests for both allelic and genotypic homogeneity. Besides the known problem of the allelic procedure, we show that whenever these tests are used, an incoherence may arise: sometimes the genotypic homogeneity hypothesis is not rejected, but the allelic hypothesis is. As we argue, this is logically impossible. Some methods that were recently proposed implicitly rely on the idea that this does not happen. In an attempt to correct this incoherence, we describe an alternative frequentist approach that is appropriate even when Hardy-Weinberg equilibrium does not hold. It is then shown that the problem remains and is intrinsic of frequentist procedures. Finally, we introduce the Full Bayesian Significance Test to test both hypotheses and prove that the incoherence cannot happen with these new tests. To illustrate this, all five tests are applied to real and simulated datasets. Using the celebrated power analysis, we show that the Bayesian method is comparable to the frequentist one and has the advantage of being coherent. CONCLUSIONS: Contrary to more traditional approaches, the Full Bayesian Significance Test for association studies provides a simple, coherent and powerful tool for detecting associations.

Optimizing the CAMCOG test in the screening for mild cognitive impairment and incipient dementia: saving time with relevant domains.

OBJECTIVE: To identify the CAMCOG sub-items that best contribute for the identification of patients with mild cognitive impairment (MCI) and incipient Alzheimer's disease (AD) in clinical practice. METHODS: Cross-sectional assessment of 272 older adults (98 MCI, 82 AD, and 92 controls) with a standardized neuropsychological battery and the CAMCOG schedule. Backward logistic regression analysis with diagnosis (MCI and controls) as dependent variable and the sub-items of the CAMCOG as independent variable was carried out to determine the CAMCOG sub-items that predicted the diagnosis of MCI. RESULTS: Lower scores on Language, Memory, Praxis, and Calculation CAMCOG sub-items were significantly associated with the diagnosis of MCI. A composite score obtained by the sum of these scores significantly discriminated MCI patients from comparison groups. This reduced version of the CAMCOG showed similar diagnostic accuracy than the original schedule for the identification of patients with MCI as compared to controls (AUC = 0.80 ± 0.03 for the reduced CAMCOG; AUC = 0.79 ± 0.03 for the original CAMCOG). CONCLUSION: This reduced version of the CAMCOG had similar diagnostic properties as the original CAMCOG and was faster and easier to administer, rendering it more suitable for the screening of subtle cognitive deficits in general clinical practice.