Search

There exists a vast amount of research on how students, in higher education, approach their studying and learning. Such research resulted in a multitude of questionnaires and tools to capture the way students perform in higher education institutions. One of these tools is the Approaches to Study Skills Inventory for Students (ASSIST) that was developed in the ’80 s and ’90 s. This inventory broadly classifies students, as approaching their study, in a deep, a strategic and/or a surface manner. Although it has gone through rigorous validation in many studies, there exist no publicly available dataset of the results of these studies and so the raw datasets cannot be re-used or integrated with other similar datasets. Here, we report and make publicly available the raw data of an 8-year longitudinal survey using this inventory in a cohort study of 1181 students from a department of a higher education institution. We validated our dataset using reliability analyses that confirmed, and compared well, with the results of previous studies., Measurement(s) Study • learning Technology Type(s) questionnaire Factor Type(s) temporal interval • age • gender • parental education level Sample Characteristic - Organism Homo sapiens Sample Characteristic - Location Athens Machine-accessible metadata file describing the reported data: 10.6084/m9.figshare.14665848

The expected-posterior prior (EPP) and the power-expected-posterior (PEP) prior are based on random imaginary observations and offer several advantages in objective Bayesian hypothesis testing. The use of sufficient statistics, when these exist, as a way to redefine the EPP and PEP prior is investigated. In this way the dimensionality of the problem can be reduced, by generating samples of sufficient statistics instead of generating full sets of imaginary data. On the theoretical side it is proved that the new EPP and PEP definitions based on imaginary training samples of sufficient statistics are equivalent with the standard definitions based on individual training samples. This equivalence provides a strong justification and generalization of the definition of both EPP and PEP prior, since from the individual samples or from the sufficient samples the criteria coincide. This avoids potential inconsistencies or paradoxes when only sufficient statistics are available. The applicability of the new definitions in different hypotheses testing problems is explored, including the case of an irregular model. Calculations are simplified; and it is shown that when testing the mean of a normal distribution the EPP and PEP prior can be expressed as a beta mixture of normal priors. The paper concludes with a discussion about the interpretation and the benefits of the proposed approach.

Power-expected-posterior (PEP) priors have been recently introduced as generalized versions of the expected-posterior-priors (EPPs) for variable selection in Gaussian linear models. They are minimally-informative priors that reduce the effect of training samples under the EPP approach, by combining ideas from the power-prior and unit-information-prior methodologies. In this paper we prove the information consistency of the PEP methodology, when using the independence Jeffreys as a baseline prior, for the variable selection problem in normal linear models.

Graphical models are used for expressing conditional independence relationships among variables by the means of graphs, whose structure is typically unknown and must be inferred by the data at hand. We propose a theoretically sound Objective Bayes procedure for graphical model selection. Our method is based on the Expected‐Posterior Prior and on the Power‐Expected‐Posterior Prior. We use as input of the proposed methodology a default improper prior and suggest computationally efficient approximations of Bayes factors and posterior odds. In a variety of simulated scenarios with varying number of nodes and sample sizes, we show that our method is highly competitive with, or better than, current benchmarks. We also discuss an application to protein‐signaling data, which wieldy confirms existing results in the scientific literature.

It is known that family influences social trends, behavior and attitude as well as smoking. We conducted a survey in 534 adolescents from schools in Athens as part of the smoking prevention program “PARE ANASA” of HTS. The smoking behavior of parents at adolescents home is portrayed in Table 1 Significant statistical dependence was found between parental smoking and the tendency of children to smoke (p The corresponding percentage given that either the father or mother is an ex-smoker is 36.6%. When both parents smoke the percentage is 46.8% and are expected to have 2.44 more likelihood of smoking (OR=2.44, 95% CI:1.14-5.25, p=0.022). When the father/mother smoke, the % is about the same 29% and when another member smokes the percentage is 56.3%. When only guests are allowed, 37% of children reported to have smoked a tobacco product, and 28% when not allowed smoking. With permitted smoking, 45% reported to have smoked, i.e it is expected to have 2.38 times (OR=2.38, 95% CI: 1.25-4.61, p=0.009) more relative likelihood to have smoked some tobacco products and it is expected to have 3.83 times greater relative likelihood to have smoked in more days, >1cig. in the last 30 days (OR=3.83,95%CI:1.27-12.47, p=0.02) compared to children with smoking prohibition at home adjusted for other factors. Thus, the effect of parental smoking is paramount to the smoking behavior of adolescents, keeping the rules to maintain a smoke free environment at home can be functionally protective. Preventive programs should include parents’ education as they are role models.

The impact of peer group concerning heath attitudes is very important in adolescents. We conducted a survey giving 534 questionnaires in students of secondary schools in 9 schools of Athens. The percentage of adolescents that responded that may smoke in the future was 76.5%, while most of all their friends are smoking, 40% when their best friend is smoking, 43.9% when some of their friends are smoking and 12.6% when no one is smoking (Fig1). There is a significant trend in adolescents willing to smoke and the smoking attitude of their friends. (p≈2.2*10-16 88.2% of adolescents have smoked or tried a tobacco product when most or more of their friends were smoking, 40% when their best friend is smoking, 50.8% when some of their friends are smoking and 13% when none is smoking (Fig2). There is a strong correlation between the smoking attitude of adolescents and those of their friends (p≈2.2*10-16 There is a significant correlation between the days that adolescents smoked more than 1 cigarette the last month and the smoking attitude of their friends. p-value=2.505*10-5 There is a significant impact of peers smoking in both the smoking attitudes and trends of adolescent. Thus, prevention programs in schools should also include the development of peer pressure resistance skills.

The problem of transformation selection is thoroughly treated from a Bayesian perspective. Several families of transformations are considered with a view to achieving normality: the Box–Cox, the Modulus, the Yeo & Johnson, and the Dual transformation. Markov chain Monte Carlo algorithms have been constructed in order to sample from the posterior distribution of the transformation parameter associated with each competing family T. We investigate different approaches to constructing compatible prior distributions for over alternative transformation families. Selection and discrimination between different transformation families are attained via posterior model probabilities. Although there is no choice of transformation family that can be universally applied to all problems, empirical evidence suggests that some particular data structures are best treated by specific transformation families. For example, skewness is associated with the Box–Cox family while fat-tailed distributions are efficiently treated using the Modulus transformation. The Canadian Journal of Statistics xx: 1–24; 2015 © 2015 Statistical Society of Canada Resume Les auteurs traitent de la selection d'une transformation des donnees dans une perspective bayesienne. Ils considerent plusieurs familles de transformations afin d'atteindre la normalite : la transformee de Box-Cox et ses generalisations attribuees a John et Draper (transformee du module), Yeo et Johnson ainsi que Yang (transformee duale). Les auteurs construisent des algorithmes Monte Carlo par chai nes de Markov afin de generer des donnees selon la distribution a posteriori du parametre associe au choix de famille T. Ils etudient differentes approches afin de construire des lois a priori compatibles permettant a de couvrir toutes les familles de transformations considerees. Ils effectuent la selection d'une famille de transformations par rapport aux autres sur la base de probabilites a posteriori. Bien qu'aucun choix de famille ne s'applique de facon universelle, les auteurs observent empiriquement que certaines transformations semblent adaptees a des structures de donnees particulieres. Par exemple, la transformee de Box-Cox est bien adaptee aux donnees asymetriques alors qu'un probleme de queues lourdes est traite efficacement par la transformee du module. La revue canadienne de statistique xx: 1–24; 2015 © 2015 Societe statistique du Canada

We provide a review of prior distributions for objective Bayesian analysis. We start by examining some foundational issues and then organize our exposition into priors for: i) estimation or prediction; ii) model selection; iii) high-dimensional models. With regard to i), we present some basic notions, and then move to more recent contributions on discrete parameter space, hierarchical models, nonparametric models, and penalizing complexity priors. Point ii) is the focus of this paper: it discusses principles for objective Bayesian model comparison, and singles out some major concepts for building priors, which are subsequently illustrated in some detail for the classic problem of variable selection in normal linear models. We also present some recent contributions in the area of objective priors on model space. With regard to point iii) we only provide a short summary of some default priors for high-dimensional models, a rapidly growing area of research.

Expected-posterior priors (EPPs) have been proved to be extremely useful for testing hypotheses on the regression coefficients of normal linear models. One of the advantages of using EPPs is that impropriety of baseline priors causes no indeterminacy in the computation of Bayes factors. However, in regression problems, they are based on one or more training samples, that could influence the resulting posterior distribution. On the other hand, the power-expected-posterior priors are minimally-informative priors that reduce the effect of training samples on the EPP approach, by combining ideas from the power-prior and unit-information-prior methodologies. In this paper, we prove the consistency of the Bayes factors when using the power-expected-posterior priors, with the independence Jeffreys as a baseline prior, for normal linear models, under very mild conditions on the design matrix.

Summary The measurement and improvement of the quality of health care are important areas of current research and development. A judgement of appropriateness of medical outcomes in hospital quality-of-care studies must depend on an assessment of patient sickness at admission to hospital. Indicators of patient sickness often must be abstracted from medical records, and some variables are more expensive to measure than others. Quality-of-care studies are frequently undertaken in an environment of cost restriction; thus any scale measuring patient sickness must simultaneously respect two optimality criteria: high predictive accuracy and low cost. Here we examine a variable selection strategy for construction of a scale of sickness in which predictive accuracy is optimized subject to a bound on cost. Conventional model search algorithms (such as those based on standard reversible jump Markov chain Monte Carlo (RJMCMC) sampling) in our setting will often fail, because of the existence of multiple modes of the criterion function with movement paths that are forbidden because of the cost restriction. We develop a population-based trans-dimensional RJMCMC (population RJMCMC) algorithm, in which ideas from the population-based MCMC and simulated tempering algorithms are combined. Comparing our method with standard RJMCMC sampling, we find that the population-based RJMCMC algorithm moves successfully and more efficiently between distant neighbourhoods of ‘good’ models, achieves convergence faster and has smaller Monte Carlo standard errors for a given amount of central processor unit time. In a case-study of n = 2532 pneumonia patients on whom p = 83 sickness indicators were measured, with marginal costs varying from smallest to largest across the predictor variables by a factor of 20, the final model chosen by population RJMCMC sampling, on the basis of both highest posterior probability and specifying the median probability model, was clinically sensible for pneumonia patients and achieved good predictive ability while capping data collection costs.

Traditional variable-selection strategies in generalized linear models (GLMs) seek to optimize a measure of predictive accuracy without regard for the cost of data collection. When the purpose of such model building is the creation of predictive scales to be used in future studies with constrained budgets, the standard approach may not be optimal. We propose a Bayesian decision-theoretic framework for variable selection in binary-outcome GLMs where the budget for data collection is constrained and potential predictors may vary considerably in cost. The method is illustrated using data from a large study of quality of hospital care in the U.S. in the 1980s. Especially when the number of available predictors p is large, it is important to use an appropriate technique for optimization (e.g., in an application presented here where p = 83, the space over which we search has 283 ≐ 1025 elements, which is too large to explore using brute force enumeration). Specifically, we investigate simulated annealing (SA), ...

In the context of the expected-posterior prior (EPP) approach to Bayesian variable selection in linear models, we combine ideas from power-prior and unit-information-prior methodologies to simultaneously (a) produce a minimally-informative prior and (b) diminish the effect of training samples. The result is that in practice our power-expected-posterior (PEP) methodology is sufficiently insensitive to the size $n^{*}$ of the training sample, due to PEP’s unit-information construction, that one may take $n^{*}$ equal to the full-data sample size $n$ and dispense with training samples altogether. This promotes stability of the resulting Bayes factors, removes the arbitrariness arising from individual training-sample selections, and greatly increases computational speed, allowing many more models to be compared within a fixed CPU budget. We find that, under an independence Jeffreys (reference) baseline prior, the asymptotics of PEP Bayes factors are equivalent to those of Schwartz’s Bayesian Information Criterion (BIC), ensuring consistency of the PEP approach to model selection. Our PEP prior, due to its unit-information structure, leads to a variable-selection procedure that — in our empirical studies — (1) is systematically more parsimonious than the basic EPP with minimal training sample, while sacrificing no desirable performance characteristics to achieve this parsimony; (2) is robust to the size of the training sample, thus enjoying the advantages described above arising from the avoidance of training samples altogether; and (3) identifies maximum-a-posteriori models that achieve better out-of-sample predictive performance than that provided by standard EPPs, the $g$ -prior, the hyper- $g$ prior, non-local priors, the Least Absolute Shrinkage and Selection Operator (LASSO) and Smoothly-Clipped Absolute Deviation (SCAD) methods.

To describe the dietary patterns of 10 European countries and their socio-demographic determinants, using the comparable between-countries DAFNE data. Analysis of standardized and postharmonized data collected through the national household budget surveys. Nationally representative surveys undertaken in 10 European countries, generally in the second half of the 1990s. The differences in the fruit and vegetable consumption previously identified between Mediterranean and Northern European countries seem to be leveling out, particularly in relation to fruit consumption. Pulses, however, still characterize the diet of the Mediterraneans. Straying from their traditional food choices, Mediterraneans recorded high availability of unprocessed red meat, while Central and Northern Europeans preferably consumed meat products. The household availability of beverages (alcoholic and non-alcoholic) is generally higher among Central and Northern European populations. Principal component (PC) analysis led to the identification of two dietary patterns in each of the 10 countries. The first was similar in all countries and indicated ‘wide-range’ food buyers. The second was slightly more varied and described ‘beverage and convenience’ food buyers. PC1 was common among households of retired and elderly members, while PC2 was common among households located in urban or semi-urban areas and among adult Scandinavians living alone. The dietary patterns identified point towards a progressive narrowing of dietary differences between North and South European countries. The comparable between-countries DAFNE data could prove useful in ecological studies, in the formulation of dietary guidelines and public health initiatives addressing specific population groups. European Commission.

Objective: To investigate the association between season of birth and psychosis, and to assess whether any association is caused by seasonal fluctuations in foetal growth or other related exposures. Method: Cohort of 747 432 Swedish males and females born between 1973 and 1980 and followed up from 16 years of age to 31 December 1999. Psychiatric admissions were identified using the Swedish Inpatient Discharge Register. The analysis is based on the 696 025 subjects with complete data. Results: A total of 506 (0.07%) subjects developed schizophrenia and 879 (0.13%) non-affective non-schizophrenic psychoses. There was a moderate increased risk of schizophrenia amongst winter births, hazard ratio 1.23 (95% confidence interval 0.96–1.59), but this did not reach conventional levels of statistical significance. There was no association with non-affective psychoses. We found no evidence that associations were confounded by measures of foetal growth or maternal socioeconomic position. There was no evidence that seasonal effects on schizophrenia differed in men and women. Conclusion: Season of birth associations with schizophrenia do not appear to be confounded by birth-related exposures.

Factors influencing fetal and childhood growth may affect a person's risk of developing schizophrenia. Associations of size at birth and body size in young adulthood with schizophrenia and other nonaffective psychoses were assessed in a cohort of 334,577 Swedish male conscripts born in 1973-1980 for whom linked birth, census, hospital admission, and adult height and weight data were available. Complete data on all study variables were available for 246,655 subjects. Over a mean 3.4-year follow-up beginning at age 18 years, 80 subjects developed schizophrenia and 124 developed other nonaffective psychoses. A reverse J-shaped association was found between gestation-adjusted birth weight and schizophrenia. The hazard ratios were 7.03 (95% confidence interval: 1.59, 31.10) for males of low birth weight (2.5 kg) and 3.37 (95% confidence interval: 1.68, 6.74) for those of high birth weight (4.0 kg). Birth weight was not strongly related to other nonaffective psychoses. Taller males had a reduced risk of psychosis. The lowest risks were seen for low birth weight males who became tall adults. The associations with birth weight indicate that fetal exposures, including possible effects of gestational diabetes, are important in the etiology of psychosis. The role of childhood exposures, as indexed by adult height and body mass index, appears to be less strong.

Background. Although urban place of birth has been identified as a risk factor for schizophrenia, the extent to which this association is mediated by socially patterned risk factors such as obstetric complications and childhood socio-economic position is unclear. The diagnostic specificity of the association within the clinical psychotic syndromes is also unclear.Method. A population cohort of 696025 males and females, born in Sweden between 1973 and 1980 and with linked birth and socio-economic data was followed up from age 16 for up to 9·8 years. Hospitalized cases of schizophrenia and other non-affective psychosis were identified from the Swedish Inpatient Discharge Register. We examined associations of these disorders with a three-level measure of urbanicity of birthplace before and after controlling for measures of foetal nutrition, obstetric complications and level of maternal education.Results. Urban compared to rural birthplace was associated both with increased risk of adult onset schizophrenia (hazard ratio 1·34, CI 0·91–1·96) and other non-affective psychoses (hazard ratio 1·63, CI 1·18–2·26). None of these associations was greatly affected by adjustment for obstetric complications or maternal educational level. In the group of other non-affective psychoses urban–rural differences in disease risk were strongest among those born in the winter months.Conclusion. Urbanization of birthplace is associated with increased risk of non-affective psychosis but this is not confined to narrowly defined cases. The magnitude of the association in Sweden is lower than that reported in other studies. Causal factors underlying this association appear to operate independently of risks associated with obstetric complications and parental educational status.

The power-conditional-expected-posterior (PCEP) prior developed for variable selection in normal regression models combines ideas from the power-prior and expected-posterior prior, relying on the concept of random imaginary data, and provides a consistent variable selection method which leads to parsimonious selection. In this paper, the PCEP methodology is extended to generalized linear models (GLMs). We define the PCEP prior in the GLM setting, explain the connections to other default model-selection priors, and present various posterior representations which can be used for model-specific posterior inference or for variable selection. The method is implemented for a logistic regression example with Bernoulli data. Results indicate that the PCEP prior leads to parsimonious selection for logistic regression models, similarly to the case of normal regression. Current limitations in generalizing the applicability of PCEP and possible solutions are discussed.

The power-expected-posterior (PEP) prior provides an objective, automatic, consistent and parsimonious model selection procedure. At the same time it resolves the conceptual and computational problems due to the use of imaginary data. Namely, (i) it dispenses with the need to select and average across all possible minimal imaginary samples, and (ii) it diminishes the effect that the imaginary data have upon the posterior distribution. These attributes allow for large sample approximations, when needed, in order to reduce the computational burden under more complex models. In this work we generalize the applicability of the PEP methodology, focusing on the framework of generalized linear models (GLMs), by introducing two new PEP definitions which are in effect applicable to any general model setting. Hyper-prior extensions for the power parameter that regulates the contribution of the imaginary data are introduced. We further study the validity of the predictive matching and of the model selection consistency, providing analytical proofs for the former and empirical evidence supporting the latter. For estimation of posterior model and inclusion probabilities we introduce a tuning-free Gibbs-based variable selection sampler. Several simulation scenarios and one real life example are considered in order to evaluate the performance of the proposed methods compared to other commonly used approaches based on mixtures of $g$ -priors. Results indicate that the GLM-PEP priors are more effective in the identification of sparse and parsimonious model formulations.

The problem motivating the paper is the quantification of students' preferences regarding teaching/coursework quality, under certain numerical restrictions, in order to build a model for identifying, assessing and monitoring the major components of the overall teaching quality. We propose a Bayesian hierarchical beta regression model, with a Dirichlet prior on the model coefficients. The coefficients of the model can then be interpreted as weights and thus they measure the relative importance that students give to the different attributes. This approach not only allows for the incorporation of informative prior when it is available but also provides user-friendly interfaces and direct probability interpretations for all quantities. Furthermore, it is a natural way to implement the usual constraints for the model coefficients. This model is applied to data collected in 2009 and 2013 from undergraduate students in the Panteion University, Athens, Greece and besides the construction of an instrument for the assessment and monitoring of teaching quality, it gave some input for a preliminary discussion on the association of the differences in students' preferences between the two time-periods with the current Greek socioeconomic transformation. Results from the proposed approach are compared with the ones obtained by two alternative statistical techniques.

We use Bayesian decision theory to address a variable selection problem arising in attempts to indirectly measure the quality of hospital care, by comparing observed mortality rates to expected values based on patient sickness at admission. Our method weighs data collection costs against predictive accuracy to find an optimal subset of the available admission sickness variables. The approach involves maximizing expected utility across possible subsets, using Monte Carlo methods based on random division of the available data into N modeling and validation splits to approximate the expectation. After exploring the geometry of the solution space, we compare a variety of stochastic optimization methods –- including genetic algorithms (GA), simulated annealing (SA), tabu search (TS), threshold acceptance (TA), and messy simulated annealing (MSA) –- on their performance in finding good subsets of variables, and we clarify the role of N in the optimization. Preliminary results indicate that TS is somewhat better than TA and SA in this problem, with MSA and GA well behind the other three methods. Sensitivity analysis reveals broad stability of our conclusions.

The problem of transformation selection is thoroughly treated from a Bayesian perspective. Several families of transformations are considered with a view to achieving normality: the Box-Cox, the Modulus, the Yeo and Johnson and the Dual transformation. Markov Chain Monte Carlo algorithms have been constructed in order to sample from the posterior distribution of the transformation parameter λ T associated with each competing family T. We investigate different approaches to constructing compatible prior distributions for λ T over alternative transformation families, using the power-prior and the unit-information prior approaches. In order to distinguish between different transformation families, posterior model probabilities have been calculated. Using simulated datasets, we show the usefulness of our approach.

The Zellner's g-prior and its recent hierarchical extensions are the most popular default prior choices in the Bayesian variable selection context. These prior setups can be expressed as power-priors with fixed set of imaginary data. In this article, we borrow ideas from the power-expected-posterior (PEP) priors to introduce, under the g-prior approach, an extra hierarchical level that accounts for the imaginary data uncertainty. For normal regression variable selection problems, the resulting power-conditional-expected-posterior (PCEP) prior is a conjugate normal-inverse gamma prior that provides a consistent variable selection procedure and gives support to more parsimonious models than the ones supported using the g-prior and the hyper-g prior for finite samples. Detailed illustrations and comparisons of the variable selection procedures using the proposed method, the g-prior, and the hyper-g prior are provided using both simulated and real data examples. Supplementary materials for this article are av...

BackgroundThe traditional Serfling-type approach for influenza-like illness surveillance requires long historical time-series. We retrospectively evaluated the use of recent, short, historical time-series for recognizing the onset of community outbreaks of respiratory tract infections (RTIs).MethodsThe data used referred to the proportion of diagnoses for upper or lower RTIs to total diagnoses for house-call visits, performed by a private network of medical specialists (SOS Doctors) in the metropolitan area of Athens, Greece, between January 01, 2000 and October 12, 2008. The reference standard classification of the observations was obtained by generating epidemic thresholds after analyzing the full 9-year period. We evaluated two different alert generating methods [simple regression and cumulative sum (CUSUM), respectively], under a range of input parameters, using data for the previous running 4-6 week period. These methods were applied if the previous weeks contained non-aberrant observations.ResultsWe found that the CUSUM model with a specific set of parameters performed marginally better than simple regression for both groups. The best results (sensitivity, specificity) for simple regression and CUSUM models for upper RTIs were (1.00, 0.82) and (0.94, 0.93) respectively. Corresponding results for lower RTIs were (1.00, 0.80) and (0.93, 0.91) respectively.ConclusionsShort-term data for house-call visits can be used rather reliably to identify respiratory tract outbreaks in the community using simple regression and CUSUM methods. Such surveillance models could be particularly useful when a large historical database is either unavailable or inaccurate and, thus, traditional methods are not optimal.

In the field of quality of health care measurement, one approach to assessing patient sickness at admission involves a logistic regression of mortality within 30 days of admission on a fairly large number of sickness indicators (on the order of 100) to construct a sickness scale, employing classical variable selection methods to find an ``optimal'' subset of 10--20 indicators. Such ``benefit-only'' methods ignore the considerable differences among the sickness indicators in cost of data collection, an issue that is crucial when admission sickness is used to drive programs (now implemented or under consideration in several countries, including the U.S. and U.K.) that attempt to identify substandard hospitals by comparing observed and expected mortality rates (given admission sickness). When both data-collection cost and accuracy of prediction of 30-day mortality are considered, a large variable-selection problem arises in which costly variables that do not predict well enough should be omitted from the final scale. In this paper (a) we develop a method for solving this problem based on posterior model odds, arising from a prior distribution that (1) accounts for the cost of each variable and (2) results in a set of posterior model probabilities that corresponds to a generalized cost-adjusted version of the Bayesian information criterion (BIC), and (b) we compare this method with a decision-theoretic cost-benefit approach based on maximizing expected utility. We use reversible-jump Markov chain Monte Carlo (RJMCMC) methods to search the model space, and we check the stability of our findings with two variants of the MCMC model composition ($\mathit{MC}^3$) algorithm., Published in at http://dx.doi.org/10.1214/08-AOAS207 the Annals of Applied Statistics (http://www.imstat.org/aoas/) by the Institute of Mathematical Statistics (http://www.imstat.org)

BackgroundImpaired intellectual performance is associated with an increased risk of schizophrenia.AimsTo investigate whether this association is due to the influence of prenatal and early childhood exposures on both intellectual development and the risk of schizophrenia.MethodCohort of 197 613 Swedish male conscripts with linked birth, census and hospital admission data together with five measures of verbal and non-verbal intellectual performance recorded at conscription.Results109 643 subjects had complete data; over a mean 5-year follow-up, 60 developed schizophrenia and 92 developed other non-affective psychoses. Poor scores for each of the five tests were associated with 3-to 14-fold increased risk of psychosis, particularly schizophrenia. Controlling for birth-related exposures, including birth weight, and parental education did not attenuate these associations.Results109 643 subjects had complete data; over amean 5-year follow-up,60 developed schizophrenia and 92 developed other non-affective psychoses. Poor scores for each of the five testswere associatedwith 3-to 14-foldincreasedrisk of psychosis, particularly schizophrenia. Controlling for birth-related exposures, including birthweight, and parental education didnot attenuate these associations.ConclusionsPoor intellectual performance at 18 years of age is associated with early-onset psychotic disorder. Associations do not appear to be confounded by prenatal adversity or childhood circumstances, as indexed by parental education.