Análise de cluster para determinação dos fatores associados às alterações da composição corporal em pacientes com doença inflamatória intestinal
Carregando...
Arquivos
Data
2019-08-20
Autores
Gondo, Fernanda Futino [UNESP]
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
A Doença Inflamatória Intestinal (DII) apresenta curso variável. O tratamento adequado da fase aguda pode levar à remissão clínica da doença, cujo seguimento ocorre ambulatorialmente. Nesta condição, os pacientes mantêm hábito de vida normal e alguns ainda permanecem com inflamação. Esta situação de doença e estilo de vida pode interferir na composição corporal e no estado nutricional dos pacientes. Doença de Crohn (DC) e Retocolite Ulcerativa (RCU) são doenças heterogêneas em diversos aspectos, dentre eles na composição corporal e no estado nutricional. O objetivo do estudo foi avaliar clusters relacionados à variação do estado nutricional na DIl. Foi realizado estudo transversal com pacientes com DII, submetidos a avaliação clínica (Crohn’s Disease Activity Index (CDAI), escore de Mayo e uso de medicações), nutricional (recordatório de 24 horas, International Physical Activity Questionnaire (IPAQ), peso, estatura, Índice de Massa Corporal (IMC), absorciometria por raios-X de dupla energia (DEXA), análise de bioimpedância elétrica (BIA), força de preensão manual, teste de caminhada de 6 minutos) e laboratorial (hemoglobina (Hb), hematócrito (Ht), Proteína C Reativa (PCR), velocidade de hemossedimentação (VHS), albumina). Com base nestes parâmetros, foram diagnosticados com Desnutrição (Global Leadership Initiative on Malnutrition, GLIM) e Sarcopenia (European Working Group on Sacopenia in Older People 2, EWGSOP2). Foram realizados testes estatísticos descritivos por meio do cálculo de média e desvio-padrão para variáveis numéricas com distribuição normal ou mediana e quartis (Q1-Q3) para as variáveis com distribuição não normal e frequências e proporções para as variáveis qualitativas. A associação entre as variáveis categóricas foi realizada pelo teste ² ou Exato de Fisher. A associação entre as variáveis contínuas foi realizada pelos testes t ou Mann-Whitney. Análise de cluster com o método de Ward (avaliar a distância entre os clusters) e método K means que possibilitam minimizar a variação intra grupos e maximizar a variação entre grupos. Para comparação entre clusters das variáveis contínuas foi utilizado ANOVA One-Way seguido de método de Holm-Sidak para comparações de distribuição normal. Para variáveis categóricas, teste de Kruskal Wallis seguido de método de Dunn’s quando a distribuição não foi normal. p≤ 0,05. Foram avaliados 64 pacientes, com idade média de 44,3 ± 11,6 anos, 73,4% do sexo feminino, e 55,1% diagnosticados com RCU. Do total, 9,4% apresentava atividade clinica da doença e 37,5% atividade endoscópica. Quanto as medicações utilizadas, os participantes faziam uso de ácido 5-aminosalicílico (5-ASA) (35,9%); Corticosteroides (12,5%); Imunossupressores (32,8%) e Terapia Biológica (17,2%). Não houve diferença significativa entre DC e RCU segundo as variáveis analisadas, exceto para o valor médio da idade (p= 0,04). A análise de cluster permitiu a subdivisão da amostra em 8 clusters, cada qual com características relativamente homogêneas, porém diferentes entre si, que possibilitou nomea-los segundo suas características gerais e relacionadas à composição corporal: I) Caquéticos; II) Sobrepesos caquéticos; III) Nutridos em fase inicial da doença; IV) Normais; V) Sobrepeso; VI: Sedentários; VII: Sobrepesos sarcopênicos; e VIII: Indeterminado obesos. A análise dos 8 clusters possibilitou a identificação de grupos heterogêneos por meio de seus mecanismos fisiopatológicos predominantes relacionados às alterações da composição corporal.
Inflammatory Bowel Disease (IBD) presents a variable course. Adequate treatment of the acute phase may lead to clinical remission of the disease, which is followed in the outpatient clinic. In this condition, patients maintain normal life habit and some still remain with inflammation. Both disease and lifestyle situation may interfere with the body composition and nutritional status of the patients. Crohn's disease (CD) and ulcerative colitis (UC) are heterogeneous diseases in several aspects, including body composition and nutritional status. The aim of the study was to evaluate clusters related to the variation of the nutritional status in IBD. A cross-sectional study was performed with IBD patients, by clinical (Crohn's Disease Activity Index (CDAI), Mayo score and medications), nutritional (24-hour recall, International Physical Activity Questionnaire (IPAQ), weight, stature, Body Mass Index (BMI), dual energy X-ray absorptiometry (DEXA), electrical bioimpedance (BIA), handgrip force, 6-minute walk test) and laboratorial evaluation (hemoglobin (Hb), hematocrit (Ht), C-Reactive Protein (CRP), erythrocyte sedimentation rate (ESR), albumin). Based on these parameters, were diagnosed Malnutrition (Global Leadership Initiative on Malnutrition (GLIM) and Sarcopenia (European Working Group on Sacopenia in Older People 2, EWGSOP2). Descriptive statistical tests were performed by mean ± standard deviation for numerical variables with normal or median distribution and quartiles (Q1-Q3) for variables with non-normal distribution and frequencies and proportions for qualitative variables. The association between the categorical variables was performed by the ² test or Fisher's Exact test. The association between the continuous variables was performed by t test or Mann-Whitney test. Cluster analysis with the Ward method (method to evaluate the distance between clusters) and K means was performed to minimize intra-group variation and to maximize variation between groups. ANOVA One-Way followed by Holm-Sidak method for comparison of normal distribution was used to comparison between clusters of continuous variables. For categorical variables, Kruskal Wallis test followed by Dunn's method for distribution was not normal. p≤0.05. Sixty-four patients were evaluated, with a mean age of 44.3 ± 11.6 years, 73.4% female, and 55.1% diagnosed with UC were evaluated. Of the total, 9.4% presented clinical activity of the disease and 37.5% had endoscopic activity. Regarding the medications used, the participants used 5-ASA (35.9%); Steroids (12.5%); Immunomodulators (32.8%) and Biologics (17.2%). There was no difference between CD and UC according to the analyzed variables, except for the age (p = 0.04). The cluster analysis allowed the subdivision of the sample into 8 clusters, each with relatively homogeneous but different characteristics, which made it possible to name them according to their general characteristics and related to the body composition: I) Cachetics; II) Cachetic Overweight; III) Nourished in the initial phase of the disease; IV) Normal; V) Overweight; VI: Sedentary; VII: Sarcopenic Overweight ; and VIII) Indeterminate obese. The analysis of the 8 clusters allowed the identification of heterogeneous groups through their predominant pathophysiological mechanisms related to changes in body composition.
Inflammatory Bowel Disease (IBD) presents a variable course. Adequate treatment of the acute phase may lead to clinical remission of the disease, which is followed in the outpatient clinic. In this condition, patients maintain normal life habit and some still remain with inflammation. Both disease and lifestyle situation may interfere with the body composition and nutritional status of the patients. Crohn's disease (CD) and ulcerative colitis (UC) are heterogeneous diseases in several aspects, including body composition and nutritional status. The aim of the study was to evaluate clusters related to the variation of the nutritional status in IBD. A cross-sectional study was performed with IBD patients, by clinical (Crohn's Disease Activity Index (CDAI), Mayo score and medications), nutritional (24-hour recall, International Physical Activity Questionnaire (IPAQ), weight, stature, Body Mass Index (BMI), dual energy X-ray absorptiometry (DEXA), electrical bioimpedance (BIA), handgrip force, 6-minute walk test) and laboratorial evaluation (hemoglobin (Hb), hematocrit (Ht), C-Reactive Protein (CRP), erythrocyte sedimentation rate (ESR), albumin). Based on these parameters, were diagnosed Malnutrition (Global Leadership Initiative on Malnutrition (GLIM) and Sarcopenia (European Working Group on Sacopenia in Older People 2, EWGSOP2). Descriptive statistical tests were performed by mean ± standard deviation for numerical variables with normal or median distribution and quartiles (Q1-Q3) for variables with non-normal distribution and frequencies and proportions for qualitative variables. The association between the categorical variables was performed by the ² test or Fisher's Exact test. The association between the continuous variables was performed by t test or Mann-Whitney test. Cluster analysis with the Ward method (method to evaluate the distance between clusters) and K means was performed to minimize intra-group variation and to maximize variation between groups. ANOVA One-Way followed by Holm-Sidak method for comparison of normal distribution was used to comparison between clusters of continuous variables. For categorical variables, Kruskal Wallis test followed by Dunn's method for distribution was not normal. p≤0.05. Sixty-four patients were evaluated, with a mean age of 44.3 ± 11.6 years, 73.4% female, and 55.1% diagnosed with UC were evaluated. Of the total, 9.4% presented clinical activity of the disease and 37.5% had endoscopic activity. Regarding the medications used, the participants used 5-ASA (35.9%); Steroids (12.5%); Immunomodulators (32.8%) and Biologics (17.2%). There was no difference between CD and UC according to the analyzed variables, except for the age (p = 0.04). The cluster analysis allowed the subdivision of the sample into 8 clusters, each with relatively homogeneous but different characteristics, which made it possible to name them according to their general characteristics and related to the body composition: I) Cachetics; II) Cachetic Overweight; III) Nourished in the initial phase of the disease; IV) Normal; V) Overweight; VI: Sedentary; VII: Sarcopenic Overweight ; and VIII) Indeterminate obese. The analysis of the 8 clusters allowed the identification of heterogeneous groups through their predominant pathophysiological mechanisms related to changes in body composition.
Descrição
Palavras-chave
composição corporal, análise por agrupamento, desnutrição, músculo esquelético, doenças inflamatórias do intestino, inflammatory bowel diseases, skeletal muscle, malnutrition, cluster analysis, body composition