IFEROA: Identification of the female specific etiology and risk groups for osteoarthritis

Osteoarthritis (OA) is by far the most frequent rheumatic disease, affects almost 10% of the population, occurs most often in the knee, and is the fastest increasing cause for disability. After menopausal age, a steep rise in knee OA-incidence is seen in women. As a result, knee OA occurs twice as frequent in women compared to men. In addition, OA is more disabling in women, and frequently co-occurs with atherosclerosis (AS) in women but not in men. This marked sex difference suggests female specific pathways that lead to OA, and possibly a female specific common pathway that leads to both OA and AS. However, at present, such sex specific pathways have never been investigated, and no sex specific risk profiles or treatment strategies are available. Therefore, in the proposed IFEROA research project we aim: 1. To identify sex specific pathways for the development of OA in order to clarify pathway differences between men and women 2. To identify the role of menopause in the female specific pathway 3. To identify the female specific common pathway involved in the development of both OA and AS 4. To develop a clinical prediction tool that identifies females at high risk of developing OA, or both OA and AS First we will use the full spectrum of data available within the complete Rotterdam Study (15000 persons from age 45 years). The cohort has regular 5-year follow-up measurements and a mean follow-up duration of 10 years. This cohort is unique in the availability of comprehensive prospective OA and AS data together with a complete set of data available on possible risk factors, including deep phenotyping of the complete musculoskeletal system, cardiovascular disease, and other major comorbidities. OA (X-ray of hip, knee, hands, spine at all follow-up measurements, joint specific pain measures, disability measures, and morning stiffness) and AS (Ultrasound carotid and coronary measures) have been assessed with validated outcome measures. In addition, there is detailed data on body composition, diet, medication use, lifestyle factors, social-economic factors and menopausal symptoms and status. The cohort already has classical metabolic markers (lipids, glucose, others), cartilage/bone specific markers, cardiovascular markers, inflammatory markers, lipidomics, extensive hormone profiles, and genetics. The data will be analyzed with sex stratified multivariable regression models for identification of specific OA factors in order to find out which factors might contribute to the unknown female specific pathway. Next, Structural Equation Modeling will be used to construct more complex models with the possibility to model how the risk factors influence and mediate each other resulting in OA, or OA and AS combined. The main joint under study is the knee, but the other joints will also be explored. These pathways will be validated in international and national replication cohorts. Further in-depth analyses of sex specific pathways will be done in a subcohort of the Rotterdam Study of 2600 persons of age 45-60 years with baseline and five year follow-up measurements, where additional state of the art MRI measurements for knee OA and AS are available with validated disease definitions, as well as scores for early features. This subcohort is of special interest due to the age range, and includes both pre- and post-menopausal measurements. Besides the deep phenotyping available in the complete Rotterdam Study, this focus cohort is specifically suited for in depth network pathway analysis due to the availability of other interesting markers (AGEs), and a large quantity of molecular layers (metabolomics and lipidomics, epigenomics, transcriptomics, microbiomics). Proteomics (Olink) will be newly assessed in this subcohort for the present proposal. In this subcohort data mining algorithms according to state-of-the-art network analysis approaches will be used for the in depth analysis followed by Structural Equation Modeling. Replication of relevant identified omics in these pathways will again be done in replication cohorts. The identification of the female specific pathways will deliver female specific modifiable risk factors and molecular targets, and possibly intermediate outcomes, for sex specific treatment. Finally, we will develop female specific easily useable clinical prediction rules in order to identify the persons at risk amenable for targeted treatment in a precision medicine approach. Future research should focus on testing and further developing these new targeted treatments based on the female specific pathways uncovered in this study. Our ultimate aim is to decrease the knowledge gap on sex differences in OA health. The proposal will gain essential knowledge necessary to achieve optimized health care for both sexes and aims specifically at decreasing the disease burden due to OA in women at least towards that in men. This proposal sets the first steps towards sex-specific treatment strategies in OA.