A number of methods have been developed to analyze longitudinal data with dropout. However, there is no uniformly accepted approach. Model performance, in terms of the bias and accuracy of the estimator, depends on the underlying missing data mechanism and it is unclear how existing methods will perform when little is known about the missing data mechanism. Here we evaluate methods for estimating change over time in longitudinal studies with informative dropout in three settings: using a linear mixed effect (LME) estimator in the presence of multiple types of dropout; proposing an update to the pattern mixture modeling (PMM) approach in the presence of imprecision in identifying informative dropouts; and utilizing this new approach in the presence of prognostic factor by dropout interaction. We demonstrate that amount of dropout, the proportion of dropout that is informative, and the variability in outcome all affect the performance of an LME estimator in data with a mixture of informative and non-informative dropout. When the amount of dropout is moderate to large (>20% overall) the potential for relative bias greater than 10% increases, especially with large variability in outcome measure, even under scenarios where only a portion of the dropouts are informative. Under conditions where LME models do not perform well, it is necessary to take the missing data mechanism into account. We develop a method that extends the PMM approach to account for uncertainty in identifying informative dropouts. In scenarios with this uncertainty, the proposed method outperformed the traditional method in terms of bias and coverage. In the presence of interaction between dropout and a prognostic factor, the LME model performed poorly, in terms of bias and coverage, in estimating prognostic factor-specific slopes and the interaction between the prognostic factor and time. The update to the PMM approach, proposed here, outperformed both the LME and traditional PMM. Our work suggests that investigators must be cautious with any analysis of data with informative dropout. We found that particular attention must be paid to the model assumptions when the missing data mechanism is not well understood.

No description available.

No

Les objectifs de cette thèse sont multiples. Le premier objectif est de comparer, aux niveaux comportemental et électrophysiologique, l'apprentissage implicite de structures linguistiques et musicales après l'écoute d'un langage artificiel chanté. Alors qu'au niveau comportemental, seule la structure linguistique semble être apprise, les résultats électrophysiologiques révèlent un effet N400 pour les deux dimensions, linguistique et musicale. Le deuxième objectif de cette thèse est d'évaluer comment cet apprentissage est influencé par l'expertise musicale. Nous avons comparé un groupe d'adultes musiciens à un groupe de non musiciens. Alors qu'au niveau comportemental les musiciens sont à peine meilleurs que les non musiciens dans les deux dimensions, les données électrophysiologiques révèlent, via des différences précoces (N1/P2) et tardives (N400), une meilleure segmentation chez les musiciens. De plus, les analyses en potentiels évoqués et en temps-fréquences des données électrophysiologiques enregistrées pendant les phases d'apprentissage révèlent que les musiciens apprennent plus rapidement que les non musiciens. Cependant, un lien de causalité quant aux effets de l'apprentissage de la musique ne peut être mis en évidence qu'en réalisant une étude longitudinale. Nous avons mené une telle étude chez des enfants de 8 ans à qui l'on a fait suivre un apprentissage de la musique ou de la peinture pendant 2 années. Les résultats comportementaux et électrophysiologiques révèlent un large bénéfice de l'apprentissage musical comparé à celui de la peinture démontrant l'importance de la musique dans l'éducation des enfants. / The aims of the present thesis were two-folded. Firstly, we wanted to compare behavioral and electrophysiological measures related to the implicit learning of linguistic and musical structures contained within an artificial sung language. While behavioral measures suggest that only the linguistic structure was learned, electrophysiological data revealed similar N400 effects in both linguistic and musical dimensions, suggesting that participants did also learn the musical structure. The second goal was to evaluate to what extent musical expertise can affect speech segmentation. At this aim, we compared a group of adult musicians to a group of nonmusicians. While behavioral data showed that musicians had marginally better performance than non musicians in both dimensions, electrophysiological data revealed, via early (N1/P2) and late (N400) differences, a better speech segmentation in musicians than in non musicians. Moreover, event-related potentials and time-frequency analyzes during learning revealed a faster and more efficient learning process in musicians. However, the only way to unambiguously claim causality between expertise and the observed effects requires a longitudinal approach. At this aim, we conducted a study with 8 year-old children who followed either music or painting lessons over a period of 2 years. Behavioral and electrophysiological data revealed a larger benefit of musical compared to painting training, bringing evidences for the importance of music in childrens' education.

BACKGROUND: Tobacco use remains the leading cause of morbidity and mortality in the US. Effective tobacco cessation aids are widely available, yet underutilized. Tobacco cessation brief interventions (BIs) increase quit rates. However, BI training has focused on conventional medical providers, overlooking other health practitioners with regular contact with tobacco users. The 2007 National Health Interview Survey found that approximately 20% of those who use provider-based complementary and alternative medicine (CAM) are tobacco users. Thus, CAM practitioners potentially represent a large, untapped community resource for promoting tobacco cessation and use of effective cessation aids. Existing BI training is not well suited for CAM practitioners' background and practice patterns, because it assumes a conventional biomedical foundation of knowledge and philosophical approaches to health, healing and the patient-practitioner relationship. There is a pressing need to develop and test the effectiveness of BI training that is both grounded in Public Health Service (PHS) Guidelines for tobacco dependence treatment and that is relevant and appropriate for CAM practitioners. METHODS/DESIGN: The CAM Reach (CAMR) intervention is a tobacco cessation BI training and office system intervention tailored specifically for chiropractors, acupuncturists and massage therapists. The CAMR study utilizes a single group one-way crossover design to examine the CAMR intervention's impact on CAM practitioners' tobacco-related practice behaviors. Primary outcomes included CAM practitioners' self-reported conduct of tobacco use screening and BIs. Secondary outcomes include tobacco using patients' readiness to quit, quit attempts, use of guideline-based treatments, and quit rates and also non-tobacco-using patients' actions to help someone else quit. DISCUSSION: CAM practitioners provide care to significant numbers of tobacco users. Their practice patterns and philosophical approaches to health and healing are well suited for providing BIs. The CAMR study is examining the impact of the CAMR intervention on practitioners' tobacco-related practice behaviors, CAM patient behaviors, and documenting factors important to the conduct of practice-based research in real-world CAM practices.

The overall aim of this study was to investigate the association between family structure in childhood and mental health problems in young adulthood. A prospective cohort study was conducted with 481,777 individuals with complete follow-up information, which was obtained from national registers in Sweden. Individuals who were living with only one biological parent at age 17 were compared with those who grew up with two parents with regard to retrieval of prescribed psychotropic drugs at age 35. The association was examined by Cox regression analyses with equal survival time for all individuals included in the analyses. The results demonstrated a higher risk for retrieval of psychotropic medicines among the individuals who grew up with only one parent, with hazard ratio of 1,21 (95%CI: 1,19-1,23). The multivariate analyses showed that a part of the association was explained by familial and individual factors, namely parents’ country of origin, area of residence, parents’ and the individual’s educational attainment, receipt of social benefits and parents’ history of psychiatric disorder. The results indicated that the increased risk of mental health problems among individuals who grew up with only one parent might be accounted for by various psychological, social and economic factors associated to parental separation.

Microbiome plays an important role in human health. The analysis of association between microbiome and clinical outcome has become an active direction in biostatistics research. Testing the microbiome effect on clinical phenotypes directly using operational taxonomic unit abundance data is a challenging problem due to the high dimensionality, non-normality and phylogenetic structure of the data. Most of the studies only focus on describing the change of microbe population that occur in patients who have the specific clinical condition. Instead, a statistical strategy utilizing distance-based or similarity-based non-parametric testing, in which a distance or similarity measure is defined between any two microbiome samples, is developed to assess association between microbiome composition and outcomes of interest. Despite the improvements, this test is still not easily interpretable and not able to adjust for potential covariates. A novel approach, kernel-based semi-parametric regression framework, is applied in evaluating the association while controlling the covariates. The framework utilizes a kernel function which is a measure of similarity between samples' microbiome compositions and characterizes the relationship between the microbiome and the outcome of interest. This kernel-based regression model, however, cannot be applied in longitudinal studies since it could not model the correlation between the repeated measurements. We proposed microbiome association exact tests (MAETs) in linear mixed model can deal with longitudinal microbiome data. MAETs can test not only the effect of overall microbiome but also the effect from specific cluster of the OTUs while controlling for others by introducing more random effects in the model. The current methods for multiple variance component testing are based on either asymptotic distribution or parametric bootstrap which require large sample size or high computational cost. The exact (R)LRT tests, an computational efficient and powerful testing methodology, was derived by Crainiceanu. Since the exact (R)LRT can only be used in testing one variance component, we proposed an approach that combines the recent development of exact (R)LRT and a strategy for simplifying linear mixed model with multiple variance components to a single case. The Monte Carlo simulation studies present correctly controlled type I error and provided superior power in testing association between microbiome and outcomes in longitudinal studies. Finally, the MAETs were applied to longitudinal pulmonary microbiome datasets to demonstrate that microbiome composition is associated with lung function and immunological outcomes. We also successfully found two interesting genera Prevotella and Veillonella which are associated with forced vital capacity.

`Active Ageing' has been promoted by the World Health Organisation (WHO) as a strategy for promoting the health and well-being of older people. Keeping active and involved in a range of activities not restricted to those associated with labour market participation may, it has been suggested, be beneficial for older people. In this research three domains of `engagement' were considered: paid work, formal involvement (i. e. activities such as voluntary work, attendance at training courses and participation in political organisations) and informal involvement (i. e. activities such as providing care and help to family, and looking after grandchildren). Using the first two waves of the Survey of Health, Ageing and Retirement in Europe (SHARE) and the English Longitudinal Study of Ageing (ELSA), this thesis investigated both the cross-sectional association between socio-economic, demographic and health-related variables and engagement at baseline, and the longitudinal association between engagement at baseline and self-rated health (SRH) and depressive symptoms at follow-up (controlling for baseline measures of health). The analysis was based on sample members aged 50- 69 at baseline in Denmark, France, Italy and England, countries selected to represent different welfare regimes. Cross-sectional findings showed that levels of engagement in paid work and formal activities varied across countries, whereas socio-economic, demographic and healthrelated characteristics were similarly associated with engagement in all countries under study. This suggested that country-specific factors, such as retirement policies, might play an important role in determining older people's level of engagement in paid work. Cross-sectional results also suggested that work and formal engagement were associated with good health, whereas -among certain subpopulations -informal activities were associated with bad health. Longitudinal analyses showed that, in all countries, respondents in paid work at baseline were more likely to improve their SRH and less likely to become depressed than those who were `inactive'. Formal and informal engagement were not significantly associated with health at follow-up. Longitudinal results and associations found, however, might have been biased by the high rates of attrition, as multiple imputation techniques and sensitivity analyses suggested. The current research study confirms that engagement in work is an important pathway to health in late life. More attention, however, should be paid to people's working lives, the quality of work and work conditions as these may influence participation in, and withdrawal from, the labour market.

This study explored how the structural aspects of a social network (that is, number of social ties, frequency of contact, as well as social participation), along with the functional aspect (social support), relate to mental health. Using data from the baseline questionnaire for the tracking cohort of participants in the Canadian Longitudinal Study on Aging, community-dwelling older adults aged 65-85 years old were studied. Cluster analysis was used to group individuals into different clusters, based on their structural social network characteristics. Six clusters were found, ranging from most socially integrated, to moderately integrated, to socially isolated. Univariate analyses indicated that as level of social integration decreased, individuals fared increasingly worse in terms of their mental health outcomes. Furthermore, a series of mediation analyses showed that social support mediated the relationship between social integration level, and mental health, an effect that was strongest for the most socially isolated individuals. / May 2016

[pt] O estudo se insere no âmbito das investigações sobre escolas eficazes. Utiliza dados de survey longitudinal, obtidos pela pesquisa GERES: Estudo da Geração Escolar 2005. A subamostra corresponde a 3.454 alunos da primeira série do Ensino Fundamental (ou seu equivalente em ciclo), e seus respectivos professores, distribuídos em 176 turmas, de 68 escolas pertencentes às redes municipal, privada e federal do Rio de Janeiro. Transitando-se entre as vertentes qualitativa e quantitativa, procurou-se medir, descrever e interpretar o valor agregado pelos professores às suas turmas, em leitura, especialmente no que diz respeito ao uso do livro didático de Língua Portuguesa. Aprofunda-se o tema, articulando-se as concepções de alfabetização no Brasil nas últimas décadas com as características do Programa Nacional do Livro Didático (PNLD) e com as formas de apropriação por parte dos professores em relação a essa política. A literatura referente à eficácia escolar e o aprofundamento do tema específico fundamentam as análises dos dados e conseqüentes conclusões. Na vertente qualitativa, foi construída uma escala de proficiência que, interpretada, resulta na descrição de sete níveis de aprendizagem, o que confere sentido pedagógico à escala. São apresentados os resultados alcançados pelos estudantes em leitura, tendo- se em consideração as particularidades das dependências administrativas às quais estão vinculados. Mostra-se como ocorreu a mobilidade de alunos entre os níveis de aprendizagem em função da utilização ou não do livro didático. Na vertente quantitativa, são implementados modelos de regressão linear que investigam o efeito do uso do livro didático, do tempo de experiência do professor com o referido livro e da menção recebida pelo livro no PNLD. Os resultados evidenciam um aumento significativo do aprendizado médio apresentado pelos alunos participantes do estudo, o que contradiz hipótese recorrente no meio educacional, segundo a qual o fracasso identificado nas séries iniciais do Ensino Fundamental teria origem no processo de alfabetização. Destacam-se, entre os achados, maior valor agregado pelos alunos que utilizaram livro didático e benefícios ainda maiores para estudantes cujos professores tinham mais de dois anos de experiência com o uso do mesmo livro. Em termos de política pública, a dissertação sugere que a avaliação de livros didáticos - e, em sentido mais geral, de programas educacionais - considere também o efeito sobre o aprendizado. / [en] This research is included in the investigation about efficaciousness schools. It uses longitudinal survey data from the GERES Project: School Generation Research 2005. The part of the sub-sample studied corresponds to 3454 students from the first year of the primary education (or its equivalent) and its respective teachers of 176 classes, from 68 private, municipal or federal schools in Rio de Janeiro. Moving through quantitative and qualitative approaches, the added value in reading by the teachers to the classes is measured, described and interpreted, specially related to the use of Portuguese textbooks. Through in depth study of this theme, the conceptions about Brazil alphabetization that have appeared in previous decades and have considered the characteristics of the Textbooks National Program (PNLD) and the appropriation of these politics by the teachers are articulated. The literacy referred to the scholar efficaciousness and the research into this specific theme are the basis for the data analysis and its conclusions. In the qualitative approach a proficiency scale was built and, when it is interpreted, results in a description of seven learning levels, proving its inherent significance of the scale to teaching. The reading results reached by the students are presented, considering the administrative dependences that they are linked to. The students mobility is presented within the learning levels according to whether they used or didn´t use textbooks. In the quantitative approach, a linear regression model was implanted, which investigated the effect of the use of textbooks, how much experience the teacher had with that book and what PNLD says about the same book. The study verified that there was a significant increase in the learning average of the students that participated in the research. This contradicted a recurrent hipothesis in the educational area that suggested that the failure in the beginning levels of the primary education has origin in the alphabetization process. Among the results, stands out that there was more added value by the students that used the textbook and much more benefits to the students whose teachers had more than two years experience with the same book. In terms of public politics, this dissertation suggests that the textbook evaluation - and, in a more general form, the educational programs - considers the effect on the learning too.