Early determinants of dysmenorrhea at the beginning of menstrual life

Abstract

Dysmenorrhea is a common condition among women of reproductive age with negative consequences for her and society as a whole. It sometimes appears many years after menarche and is associated with specific diseases such as endometriosis and adenomyosis. However, it often appears early in women's menstrual life and is not associated with any other macroscopically determined condition. In this situation, most likely dysmenorrhea (so-called primary) is the disease in question. Surprisingly, preliminary studies have already shown that primary dysmenorrhea translates into a clinical phenotype indicative of severe risk for the development of other chronic pains in the future. Thus, if we wish to curb and devise risk minimization strategies, there is a relevant need to identify the determining factors of its early appearance. Objectives: Therefore, the main objective of this study is to identify any factors present in childhood that increase the risk of dysmenorrhea in early menstrual life, and to develop an accurate tool for early prediction of dysmenorrhea in early menstrual life. Methods: this retrospective longitudinal study will evaluate data from ALSPAC (The Avon Longitudinal Study of Parents and Children). The main outcome will be the presence of dysmenorrhea detected in the first years of menstrual life. The age at menarche will be determined in the questionnaires applied in late childhood, adolescence and transition to adult life. The presence of menstruation and dysmenorrhea will be evaluated in the following two annual questionnaires. The universe of variables to be explored has 88,989 attributes, including information on simple nucleotide polymorphisms. The computational infrastructure for the analysis will be provided by the host institution and we will use Hadoop for massive parallel processing and guarantee of efficiency. After a rigorous process of pre-processing and exploratory data analysis, we will proceed with dimensionality reduction, followed by multivariate logistic analysis and multicollinearity analysis to identify independent association variables. In the next step, we will build several prediction models and evaluate individual and group performance. The study will be carried out at the Oxford Endometriosis CaRe Center at the University of Oxford. (AU)