DIFFERENCES IN SOME DIABETIC INDICATORS AMONG PATIENTS WITH SEXUAL DYSFUNCTION

Abstract

Erectile dysfunction is defined as the continuous inability of a man to attain a satisfactory erection for a successful sexual intercourse. Female sexual dysfunction is also defined as the inability of a woman to enjoy a complete sexual activity. A twoway between groups MANOVA was performed to investigate whether there is a significant difference between male and female, and married and single. Five dependent variables were used: age, creatinine levels, duration of diabetes, glucose levels and pulse rates. The independent variables were sex and marital status. Preliminary assumption testing was conducted to check for normality, linearity, univariate and multivariate outliers, homogeneity of variance-covariance matrices, and multicollinearity, with no serious violations noted. There was a significant difference between males and females on the dependent variables: F (5, 223) = 34.00, P = 0.00; Wilk’s lambda = 0.55. When the results for the dependent variables were considered separately using the logit and the probit regression models, only age and creatinine levels reached statistical significance. An inspection of the mean scores indicated that males recorded slightly higher scores in the dependent variables than females and the single recorded slightly higher values than the married. There was also a significant difference between married and single on the dependent variables; F (5, 223) = 25.74, P = 0.00; Wilk’s lambda = 0.63. The interaction effect was not statistically significant; F (5, 223) = 17.55, P = 0.06 and Wilk’s lambda = 0.72. Probit and logit models were used to classify and predict observations. It was revealed that logit model has a higher predictive and classification power than the probit model, though they both produced very similar results. Bootstrapping was performed to check the sensitivity of the models. It was revealed that the bootstrapped estimates converged to the actual estimates.