Logistic regression analysis of finding associated factors to predict loss weight adults in Erbil City (2018)

Abstract

One of the threatened risk factors which leads to numerous health issue is overweight or obesity. Many studies have been carried out about this problem and yet not exact cause have been found. Numerous factors are highlighted as primary ones such as, not doing exercise, unorganized daily meal, medical condition and etc. In addition, statistical analysis plays an essential role in finding the most effective factors linked to reducing body weight and since the values of response variables lies under two levels which shows no linear relationship between the outcome and explanatory variables, thus Binary Logistic Regression a family of Generalized Linear Model was performed. Logistic regression analysis is a very common tool and serves great part in health science due to the fact that most of the phenomena’s outcome have only two values (alive/dead, exposed/not exposed, presence/absence, and etc.). The overall adults who underwent losing weight and succeeded was 57.7% and 42.3% who failed. Frequently medical visits and exercise were highly significant and odds of one-unit increase, has about 3 times more chance to lose. Gender, eating-out, overeating and irregular eating were all highly significant. However, diet and number of healthy meals were not found to be associated.