The paper proposes a parametric approach to estimate the Lorenz curve and the Gini index in the context of describing exposure-disease association. Nonparametric bootstrap statistical inference method is employed for generating estimates of statistical variability for the Gini index. The index describes the overall degree of risk variation in a population, it does not indicate where in the distribution the variation may be occurring. To remedy this limitation, analysis based on the Gini index is interpreted in conjunction with percentile estimates and a measure of skewness of the Lorenz curve. To demonstrate the proposed methodology, international data on AIDS incidence for selected countries is used. Results obtained using the Lorenz-Gini methodology for estimating disease risk are compared with results obtained from an alternative approach utilizing the negative binomial regression.