Estrogen-related and other disease diagnoses preceding Parkinson’s disease

Jeanne C Latourelle^1,2, Merete Dybdahl³, Anita L Destefano^1,4, Richard H Myers¹, Timothy L Lash^2,3

¹Department of Neurology, Boston University School of Medicine, Boston MA, USA; ²Department of Epidemiology, Boston University School of Public Health, Boston MA, USA; ³Department of Clinical Epidemiology, Aarhus University Hospital, Aarhus, Denmark; ⁴Department of Biostatistics, Boston University School of Public Health, Boston MA, USA

Purpose: Estrogen exposure has been associated with the occurrence of Parkinson’s disease (PD), as well as many other disorders, and yet the mechanisms underlying these relations are often unknown. While it is likely that estrogen exposure modifies the risk of various diseases through many different mechanisms, some estrogen-related disease processes might work in similar manners and result in association between the diseases. Indeed, the association between diseases need not be due only to estrogen-related factors, but due to similar disease processes from a variety of mechanisms.

Patients and methods: All female Parkinson’s disease cases between 1982 and 2007 (n = 12,093) were identified from the Danish National Registry of Patients, along with 10 controls matched by years of birth and enrollment. Conditional logistic regressions (CLR) were used to calculate risk of PD after diagnosis of the estrogen-related diseases, endometriosis and osteoporosis, conditioning on years of birth and enrollment. To identify novel associations between PD and any other preceding conditions, CLR was also used to calculate the odds ratios (ORs) for risk of PD for 202 different categories of preceding disease diagnoses. Empirical Bayes methods were used to identify the robust associations from the over 200 associations produced by this analysis.

Results: We found a positive association between osteoporosis and osteoporotic fractures and PD (OR = 1.18, 95% confidence interval [CI] of 1.08–1.28), while a lack of association was observed between endometriosis and PD (OR = 1.37, 95% CI 0.99–1.90). Using empirical Bayes analyses, 24 additional categories of diseases, likely unrelated to estrogen exposure, were also identified as potentially associated with PD.

Conclusion: We identified several novel associations, which may provide insight into common causal mechanisms between the diseases or greater understanding of potential early preclinical signs of PD. In particular, the associations with several categories of mental disorders suggest that these may be early warning signs of PD onset or these diseases (or the causes of these diseases) may predispose to PD.

Keywords: Parkinson’s disease, estrogen, osteoporosis, endometriosis, empirical bayes