Risk factors of Parkinson disease among women from the E3N cohort – PARKIN-WOMEN

Parkinson's disease (PD) is the second more frequent neurodegenerative disease after Alzheimer’s disease. It is about 1.5 times more frequent in men than women who are exposed to estrogens throughout their life, from puberty to menopause, and eventually after through postmenopausal hormone therapy (HT) use and estrogens synthetized in adipose tissues. These observations, together with experimental data, led to the hypothesis that estrogens could play a protective role for PD. However, few cohorts, especially in Europe, have assembled a sufficient number of cases in women, to address this question through a prospective design and with validated PD cases and adequate statistical power, and their results are inconsistent. The Etude Epidémiologique des femmes de l’Education Nationale includes 98,995 women affiliated with the MGEN (Mutuelle générale de l'Éducation nationale) who have been followed for over 25 years and offers a unique opportunity to investigate risk/protective factors assessed in midlife in women. Using these data, we aim to study the association of PD incidence with lifetime hormonal exposure, including endogenous estrogens, reflected by characteristics of reproductive life, and exogenous hormones (hormonal contraceptives and postmenopausal HT).
PD cases will be validated using a three-phase procedure based on medical records and expert consensus. First, we will identify potential PD cases based on self-report by women during the follow-up and/or use of antiparkinsonian medications from drugs claim databases from MGEN. We will then contact neurologists, general practitioners, and hospitals in order to obtain information on diagnosis, symptoms, age at onset and treatments. Finally, an expert committee will validate the cases. For women for whom we will not be able to collect sufficient information for the PD classification, we will use a validated algorithm with high sensitivity and specificity to identify PD cases. At the present date, we have identified 929 incident cases until 2014, and we will pursue the identification and validation of PD cases with more recent data.
Hormonal exposure will be defined as i) exposure to endogenous estrogens: age at menarche, number of and age at pregnancies, duration of breastfeeding, age at and type of menopause, fertility life length; ii) exogenous hormonal exposure: hormonal contraceptives and postmenopausal HT, including molecules, doses, route of administration, duration of use, type of progestogens. We will take into account adiposity, assessed by the body-mass index updated at each questionnaire, and effect modification by age, smoking, and caffeine intake.
Statistical analysis will rely on Cox proportional hazard models and joint models with time dependent variables and analyses of trajectories. We will also study mortality related to PD and the effect of hormonal characteristics on mortality.
In order to strengthen causal inference, we will use Mendelian randomization using COURAGE-PD data. We will identify genetic polymorphisms associated with phenotypes such as age at menarche, age at menopause and endogenous estrogens levels and we will study their association with PD.
The strengths of this project include a prospective design with a long follow-up that allows to limit reverse causation and detection bias, a careful validation of PD cases to limit misclassification, and good statistical power to detect small effects. In addition, the Mendelian randomization approach will help to examine causality.
Through this project, the largest cohort study of PD in women to date, our objective is to improve our understanding of the aetiology of PD and to help identify specific groups at higher/lower risk several years before PD onset that could be targeted by preventive measures.