Clustering of health risk behaviors in pregnant individuals: Data from the Brazilian risk factor surveillance system for non-communicable chronic diseases.

OBJECTIVE: To investigate the clustering of health risk behaviors (HRB) and its association with demographics, physical exercise, overweight, perception of health, and diseases in Brazilian pregnant people. STUDY DESIGN: This is a cross-sectional study using data from the Risk Factor Surveillance System for Non-communicable Chronic Diseases by Telephone Survey (VIGITEL), the main health survey in Brazil. METHODS: We used data on fruit and vegetable consumption, TV time, tobacco, and alcohol abuse in individuals who reported being pregnant (n = 4553). We used latent class analysis to identify optimal HRB clustering among participants. Multinomial regression (odds ratio [OR] and 95% confidence intervals [95%CI]) was applied to identify factors associated with HRB cluster. RESULTS: Three clustering classes were identified: "without HRB cluster" (i.e., least unhealthy behaviors) (n = 2402, 52,8%), "moderate HRB cluster" (n = 1983, 43,5%), and "high HRB cluster" (i.e., most unhealthy behaviors) (n = 168, 3,7%). Pregnant people aged 35-50 years (OR = 1.89, 95%CI = 1.01; 3.52) who did not practice physical exercise (OR = 1.94, 95%CI 1.11; 3.39) were more likely to be classified as "high HRB cluster". Participants with 9-11 years (OR = 0.11, 95%CI = 0.07; 0.17) and ≥ 12 (OR = 0.05, 95%CI = 0.02; 0.11) years of education had a lower likelihood of being in the "high HRB cluster". CONCLUSION: Three HRB clustering patterns were found in this study. Greater maternal age, low education, and absence of physical exercises increased the chances of being in the high HRB cluster group. Participants with higher educational levels were less likely to be in the High HRB cluster.

Depressive Profiles among Older Puerto Rican Adults.

OBJECTIVE: This study examines the heterogeneity of depressive symptoms among older adults residing on the Island of Puerto Rico and their association with risk and protective health factors. METHODS: Data from the Puerto Rican Elderly: Health Conditions study (PREHCO) to identify risk and protective factors associated with profiles of depressive symptoms. The sample was comprised of 3,114 Puerto Rican older adults ages 60 to 102 (71 ± 8.2 SD). The Geriatric Depression Scale 15-item short form (GDS-15) was used to assess the presence of depressive symptoms. Latent class analysis (LCA) was used to categorize depressive symptom classes. The association between depressive symptom classes and health status, lifestyle, health behavior and demographic characteristics were evaluated. RESULTS: A three-class model was identified, they were defined as "low depressive symptoms" (56%), "medium depressive symptoms" (31%), and "high depressive symptoms" (13%). Having chronic illnesses, mainly diabetes, heart disease and high blood pressure as well as a history of smoking, and poor self-rated vision were all associated with being in a higher depressive class. Exercising regularly and participating in volunteer activities were protective to being in a high depressive class. CONCLUSION: Chronic illnesses and poor self-rated vision were associated with being in a higher depressive classification. These findings can inform targeted interventions for sub-groups of community dwelling older adults impacted by depression.

Extracting Spurious Latent Classes in Growth Mixture Modeling With Nonnormal Errors.

Growth mixture modeling is generally used for two purposes: (1) to identify mixtures of normal subgroups and (2) to approximate oddly shaped distributions by a mixture of normal components. Often in applied research this methodology is applied to both of these situations indistinctly: using the same fit statistics and likelihood ratio tests. This can lead to the overextraction of latent classes and the attribution of substantive meaning to these spurious classes. The goals of this study are (1) to explore the performance of the Bayesian information criterion, sample-adjusted BIC, and bootstrap likelihood ratio test in growth mixture modeling analysis with nonnormal distributed outcome variables and (2) to examine the effects of nonnormal time invariant covariates in the estimation of the number of latent classes when outcome variables are normally distributed. For both of these goals, we will include nonnormal conditions not considered previously in the literature. Two simulation studies were conducted. Results show that spurious classes may be selected and optimal solutions obtained in the data analysis when the population departs from normality even when the nonnormality is only present in time invariant covariates.

Los modelos de mezcla de la Teoría de respuesta al ítem

Se presentan los modelos de mezcla de la Teoría de respuesta al ítem y cómo estos pueden ser utilizados para identificar subagrupaciones no observadas de examinados denominadas como clases latentes. Asimismo se ejemplifica la utilidad del modelo de mezcla de la Teoría de respuesta al ítem de dos parámetros, mediante el cual se logra detectar la presencia de ítems en una escala de depresión que miden de manera distinta a los examinados en comparación con el resto de ítems que componen dicho instrumento. Finalmente, se presentan algunas recomendaciones generales en cuanto a la complementariedad que debe existir entre la teoría sustantiva que subyace al desarrollo de una escala y la aplicación de los modelos de mezcla de la Teoría de respuesta al ítem.

The Item Response Theory mixture models and how these can be used to identify unobserved sub-groups of examinees, known as latent classes, are presented. The usefulness of the two-parameter mixture model parameters is exemplified by detecting the presence of items in a depression scale that measure the examinees differently in comparison with the rest of items. Finally, some general recommendations regarding the complementarity that should exist between the substantive theory underlying the development of a scale and the use of these models are given.

ELF (Enhanced Liver Fibrosis) como marcador não invasivo de fibrose hepática na hepatite C crônica / ELF (Enhanced Liver Fibrosis) as a non invasive predictor of liver fibrosis in hepatitis C

A fibrose hepática é o aspecto mais relevante e o mais importante determinante de morbimortalidade na hepatite C crônica (HCC). Historicamente, a biópsia hepática é o método de referência para avaliação da fibrose causada pela HCC, apesar de apresentar limitações. O estudo de marcadores não invasivos, que possam obviar a necessidade da biópsia, é uma área de constante interesse na hepatologia. Idealmente, a avaliação da fibrose hepática deveria ser acurada, simples, prontamente disponível, de baixo custo e informar sobre o prognóstico da patologia. Os marcadores não invasivos mais estudados são a elastografia hepática transitória (EHT) e os laboratoriais. A EHT já foi extensamente validada na HCC e está inserida na rotina de avaliação destes pacientes. Dentre os laboratoriais, existem diversos testes em continua experimentação e, até o momento, nenhum foi integrado à prática clínica no Brasil, embora já aplicados rotineiramente em outros países. O Enhanced Liver Fibrosis (ELF), um teste que dosa no soro ácido hialurônico, pró-peptídeo amino-terminal do colágeno tipo III e inibidor tissular da metaloproteinase 1, tem se mostrado bastante eficaz na detecção de fibrose hepática significativa e de cirrose na HCC. Neste estudo o ELF teve o seu desempenho avaliado em relação a biópsia hepática e demonstrou apresentar boa acurácia na detecção tanto de fibrose significativa quanto de cirrose. Na comparação com a EHT apresentou acurácia semelhante para estes mesmos desfechos, com significância estatística. No entanto, foi observada uma superestimação da fibrose com a utilização dos pontos de corte propostos pelo fabricante. Este achado está em acordo com a literatura, onde não há consenso sobre o melhor ponto de corte a ser empregado na prática clínica. Com a ampliação da casuística foi possível propor novos pontos de corte, através da análise clássica, com a biópsia hepática como padrão ouro...

Liver fibrosis is the most relevant issue concerning chronic hepatitis C (CHC) and determines its prognosis. Historically, liver biopsy has been the reference method for evaluating fibrosis related to CHC, though it presents many drawbacks. There is a continuing interest in the development of non invasive markers capable of replacing liver biopsy. The ideal surrogate for fibrosis evaluation should be accurate, simple, low cost and yield prognostic information. So far, the most well known non invasive methods are transient hepatic elastography (TE) and laboratory panels. TE has already been extensively validated and is integrated in patients routine. There is plenty of laboratory panels in continuing evaluation and some are already integrated in daily practice abroad. In Brasil, until the present moment, it is not a reality. Enhanced Liver Fibrosis (ELF) panel comprises the serum concentration of hyaluronic acid, tissue inhibitor of matrix metalloproteinases-1, and aminoterminal propeptide of type III procollagen and has demonstrated good performance in detecting significant fibrosis and cirrhosis in CHC patients. In the present study ELF had it’s performance evaluated against liver biopsy and obtained satisfactory accuracy in detecting significant fibrosis and cirrhosis. In comparison to TE no statistically significant diference was observed, for the same endpoints mentioned before. However, the application of manufacturer’s cutoff points produced overestimation of fibrosis stages. These findings are in accordance with other author’s results, in that there is no consensus so far on the most adequate cutoff points for main clinical end points. Enlarging the data permited calculating new cutoff points, through the classical statistical approach, using liver biopsy as the gold standard...