Multivariate Analysis of Data on Migraine Treatment


  • Agostino Tarsitano Dipartimento di Economia, Statistica e Finanza, Università della Calabria, Via Pietro Bucci, Cubo 1c, 87036 Rende (CS), Italy
  • Ilaria L. Amerise Dipartimento di Economia, Statistica e Finanza, Università della Calabria, Via Pietro Bucci, Cubo 1c, 87036 Rende (CS), Italy



Kostecki-Dillon, General dissimilarity coefficient, Cluster analysis, Multi-dimensional scaling.


Migraineur constitutes a multidimensional model of health disorder involving a complex combination of genetic, psychological, demographic, enviromental and economic factors. This model provides a framework to describe limitations of an individual functional ability and quality of life, and to aid in the elaboration of more adequate intervention programs for each patient. Our primary objective in this paper is a data-driven profiling of patients.

The approach followed consists of examining affinity/dissimilarity between sufferers on the basis of different family of indicators and then aggregating multiple partial matrices, where each matrix expresses a particular notion of the dissimilarity of one patient from another. One important particularity of our method is the notion of multi-dimensional dissimilarity for static and dynamic indicators, without ignoring any portion of data.

The partial dissimilarity matrices are assembled in the form of a global matrix, which is used as input of subsequent calculations, such as multidimensional scaling and cluster analysis. Our main contribution is to show how multi-scale, cross-section and longitudinal data from individuals involved in a migraine treatment program may optimally be combined to allow profiling migraine-affected patients.


Kostecki-Dillon T, Monette G, Wong P. Pine trees, comas and migraines. Newsletter. York University Institute for Social Research 2018; 14: 1-4.

Giorgino T. Computing and visualizing dynamic time warping alignments in R: the dtw package. Journal of Statistical Software 2009; 31: 1-24. DOI:

Amerise IL, Tarsitano A. Combining dissimilarity matrices by using rank correlations. Computational Statistics 2016; 31: 353-367. DOI:

Lin KY. An elementary proof of the Perron-Frobenius theorem for non-negative symmetric matrices. Chinese Journal of Phisics 1977; 15: 283-285.

Schoenberg IJ. Metric spaces and positive definite functions. Transactions of the American Mathematical Society 1938; 44: 522-536. DOI:

Gower JC. Some distance properties of latent root and vector methods used in multivariate analysis. Biometrika 1966; 5: 325-338. DOI:

Zhang F, Zhang Q. Eigenvalue inequalities for matrix product. IEEE Transactions on Automatic Control 2006; 51: 1506-1509. DOI:

Mardia KV. Some properties of classical multi-dimesional scaling. Communications in Statistics-Theory and Methods 1978; 13: 1233-1241. DOI:

Bénasséni J, Dosse MB, Joly S. On a general transformation making a dissimilarity matrix Euclidean. Journal of Classification 2007; 24: 33-51. DOI:

Caillez F, Kuntz P. A contribution to the study of the metric and Euclidean structures of dissimilarities. Psychometrika 1996; 61: 24-253. DOI:

Gower JC. Euclidean distance geometry. The Mathematical Scientist 1982; 7: 1-14.

Charrad M, Ghazzali N, Boiteau V, Niknafs A. NbClust: An R package for determining the relevant number of clusters in a data set. Journal of Statistical Software 2014; 61: 1-36. DOI:

Cox TF, Cox MAA. Multidimensional Scaling. 2nd Edition. Chaoman & Hall, Boca Raton FL, USA 2001. DOI:

Lingoes JC. Some boundary conditions for a monotone analysis of symmetric matrices. Psychometrika 1971; 36: 195-203. DOI:

Rivas Moya T. Calculating isotonic regression of the distance function in nonmetric multidimensional scaling model. Methods of Psychological Research Online 2000; 5.




How to Cite

Tarsitano, A., & Amerise, I. L. (2019). Multivariate Analysis of Data on Migraine Treatment. International Journal of Statistics in Medical Research, 8, 40–50.



General Articles