The Göttingen minipig model of obesity is used in pre-clinical research to predict clinical outcome of new treatments for metabolic diseases. However, treatment effects often remain unnoticed when using single parameter statistical comparisons due to the small numbers of animals giving rise to large variation and insufficient statistical power. The purpose of this study was to perform a correlation matrix analysis of multiple multi-scale parameters describing co-segregation of traits in order to identify differences between lean and obese minipigs. More than 40 parameters, ranging from physical, cardiovascular, inflammatory and metabolic markers were measured in lean and obese animals. Correlation matrix analysis was performed using permutation test and bootstrapping at different levels of significance. Single parameter comparisons yielded significant differences between lean and obese animals mainly for known physical traits. On the other hand, functional network analysis revealed new co-segregations, particularly in the domain of inflammatory and oxidative stress markers in the obese animals that were not present in the lean. Functional networks of lean or obese minipigs could be utilised to assess drug effects and predict changes in parameters with a certain degree of precision, on the basis of the networks confidence intervals. Comparison of functional networks in minipigs with those of human clinical data may be used to identify common parameters or co-segregations related to obesity between animal models and man.