Síndrome da fadiga crônica e absenteísmo: estudo de trabalhadores em turnos comparando stepwise e elastic-net
Arquivos
Data
2020-02-28
Autores
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Federal de Viçosa
Resumo
Caracterizada por fadiga persistente, dor muscular, dificuldades cognitivas e de sono, a Sídrome da Fadiga Crônica (CFS) tem se tornado comum nas práticas clínicas nas últimas décadas desde sua recente definição, em 1988. Estudos resultantes da contínua busca por fatores relacionados à CFS citam, dentre outros: sono irregular/insatisfatório, estresse psicológico, disfunção hormonal, deficiência de nutrientes, disfunção imunológica e infecções. Em condições de trabalho de risco o desenvolvimento da CFS pode aumentar a chance de acidentes fatais, tal como o trabalho em turnos na área de mineração que naturalmente já possui fatores evidentemente relacionados à CFS. Estudos indicam que indivíduos com má qualidade de sono e ciclos circadianos irregulares têm risco elevado de CFS, neuroticismo e absenteísmo. Uma vez que modelagem preditiva pode se mostrar efetiva tanto na pre- venção da fadiga quanto na detecção de fatores, este estudo tem o objetivo de utilizar de regressão logística ajustada por meio de dois métodos de seleção/regularização (Stepwise e Elastic-Net) para procurar modelo que descreva a relação entre variáveis bioquímicas e antropométricas com o absenteísmo. Desta forma, por meio do absenteísmo e utilizando de efeitos encontrados na bibliografia, o objetivo é procurar evidência de relação entre a CFS e absenteísmo. Os resultados obtidos mostram indícios de relação do colesterol total, HDL, LDL e Triglicerídeos com o risco de absenteísmo, relação também presente para as variáveis de sódio e potássio. Com exceção ao potássio, todas as variáveis também possuem relação similar com a CFS, de acordo com a literatura. PALAVRAS-CHAVE: Síndrome da Fadiga Crônica. Biometria. Regressão Logística. Elastic-Net.
Characterized by persistent fatigue, muscle pain, cognitive impairment and sleep difficulties, Chronic Fatigue Syndrome (CFS) has become common in clinical practices in recent decades due to its recent definition, in 1988. Studies resulting from the continuous search for related factors to CFS mention, among others: irregular/unsatisfactory sleep, psychological stress, hormonal dysfunction, nutrient deficiency, immunological dysfunction and infections. In risky working conditions, the development of CFS might increase the chance of fatal accidents, such as shift work in mines, which presents naturally some factors that are related to CFS. Studies indicate that individuals with poor sleep quality and irregular circadian cycles are at high risk for CFS, neuroticism and absenteeism. Since predictive modeling can be effective both in preventing fatigue and detecting its factors, this study aims to use logistic regression using two selection/regularization methods (Stepwise and Elastic-Net) to look for a model that describes the relationship between biochemical and anthropometric variables with absenteeism. Thus, through absenteeism and using effects found in the bibliography, the objective is to look for evidence of a relationship between CFS and absenteeism. Results show evidence of relationship between total cholesterol, HDL, LDL and triglycerides with the risk of absenteeism, relation which is also present for the sodium and potassium variables. With the exception of potassium, all variables had similar relationship with CFS, according to the literature. KEYWORDS: Chronic Fatigue Syndrome. Biometrics. Logistic Regression. Elastic-Net.
Characterized by persistent fatigue, muscle pain, cognitive impairment and sleep difficulties, Chronic Fatigue Syndrome (CFS) has become common in clinical practices in recent decades due to its recent definition, in 1988. Studies resulting from the continuous search for related factors to CFS mention, among others: irregular/unsatisfactory sleep, psychological stress, hormonal dysfunction, nutrient deficiency, immunological dysfunction and infections. In risky working conditions, the development of CFS might increase the chance of fatal accidents, such as shift work in mines, which presents naturally some factors that are related to CFS. Studies indicate that individuals with poor sleep quality and irregular circadian cycles are at high risk for CFS, neuroticism and absenteeism. Since predictive modeling can be effective both in preventing fatigue and detecting its factors, this study aims to use logistic regression using two selection/regularization methods (Stepwise and Elastic-Net) to look for a model that describes the relationship between biochemical and anthropometric variables with absenteeism. Thus, through absenteeism and using effects found in the bibliography, the objective is to look for evidence of a relationship between CFS and absenteeism. Results show evidence of relationship between total cholesterol, HDL, LDL and triglycerides with the risk of absenteeism, relation which is also present for the sodium and potassium variables. With the exception of potassium, all variables had similar relationship with CFS, according to the literature. KEYWORDS: Chronic Fatigue Syndrome. Biometrics. Logistic Regression. Elastic-Net.
Descrição
Palavras-chave
Síndrome da Fadiga Crônica, Biometria, Análise de regressão logística
Citação
NEISSE, Anderson Cristiano. Síndrome da fadiga crônica e absenteísmo: estudo de trabalhadores em turnos comparando stepwise e elastic-net. 2020. 56 f. Dissertação (Mestrado em Estatística Aplicada e Biometria) - Universidade Federal de Viçosa, Viçosa. 2020.