Regressão logística ordinal em estudos epidemiológicos

Abstract

Ordinal logistic regression models have been developed for analysis of epidemiological studies. However, the adequacy of such models for adjustment has so far received little attention. In this article, we reviewed the most important ordinal regression models and common approaches used to verify goodness-of-fi t, using R or Stata programs. We performed formal and graphical analyses to compare ordinal models using data sets on health conditions from the National Health and Nutrition Examination Survey (NHANES II).

The place where you live and self-rated health in a large urban area

“…To assess the effect of the explanatory variables on self-rated health, ordered logistic regression, also known as proportional odds model, was used 25 . This regression model was selected because of the continuous nature of the original response variable, which had been grouped into ordinal categories, and assumed the proportional odds among these different categories 25 .…”

“…This regression model was selected because of the continuous nature of the original response variable, which had been grouped into ordinal categories, and assumed the proportional odds among these different categories 25 . The proportional odds assumption was assessed by means of the approximate likelihood-ratio test for ordinal response models proposed by Wolfe & Gould 26 .…”

Simultaneidade de fatores de risco comportamentais para doença cardiovascular em estudantes universitários

“…This form of analysis is still scarce in the public health field 27 , which demonstrates the innovative and differentiated character of this research, since we could lose information when categorizing outcomes. By choosing this form of analysis, one gains in sensitivity and power, in addition to generating a single measure of association that expresses a linear effect between exposure and outcome.…”

Consumption of sugar-rich food products among Brazilian students:National School Health Survey (PeNSE 2012)

“…When POM is employed, it is assumed that OR are similar for all the categories under comparison, which corresponds to the proportional odds assumption. To interpret the model, we compared values smaller or equal to a certain category of outcome variable in relation to higher values 21 . All statistical analyses were performed using Stata software (version 12.1; StataCorp LP, College Station, USA) and commands were performed taking into account the complex sample design of PeNSE.…”