Secondary analysis of association between ethnicity and endometriosis when adjusted for age at first pregnancy using data from the PCLO-dataset
Literature reports that endometriosis is under-diagnosed as laparoscopy is the only way of verily confirming endometriosis. Various studies give estimates that it affects between 1 in 10 and 1 in 30 women. Multifactorial causes for endometriosis are suspected and the exact disease pathway is still unknown. Current findings suspect a link with hormonal dysfunction.
There is a need to know more about endometriosis as it affects a significant group of women and the severity of the symptoms of (undiagnosed) endometriosis on women's future health and well being and their pregnancies can be substantial.
Studies using data from the Nurses' Health Study have established an association between ethnicity and endometriosis. In short, Asian women have higher risk than white women, black women and Latin American women have lower risk than white women. But data is suspected to be biased as black and Latin American women have less access to laproscopic surgery which is the standard for diagnosing endometriosis. However, a study in Houston from 1985-1986 did not find any association with race (Kirshon et al, 1989)
Furthermore, CDC reports that women have postponed their pregnancies in the last decades and that Asian women have the highest age at first pregnancy (NIH, 2014). The association between endometriosis and ethnicity may therefore be confounded or modified by age at first pregnancy.
This is the reason for testing the hypothesis that women with early pregnancies have reduced risk in developing endometriosis.
Results of study could be used to help others to form hypotheses about the disease pathway of endometriosis, inform healthcare practitioners when making the assessment whether the diagnosis is likely to be endometriosis, as well as supporting women in making informed decisions about (the timing of) their pregnancies.
I need basic dataset of endometriosis cases and datasets of twice the amount of controls. I would like to match controls on age and social-economic status. The sampling ratio of 1:2 is chosen to reduce risk of low numbers of respondents in various strata.
I would like to access additional data for the cases and controls on diet, to account for (potential confounding) effects from alcohol, vegetable & fruit-, red meat- and sugar consumption. And data from the medication questionnaire to account for (hormonal) medication use; particularly oral contraceptives.
Dietary Questionnaire (DQX)
Diet History Questionnaire (DHQ)
Medication Use Questionnaire (MUQ)
I will use matched analysis, i.e. conditional logistic regression models to estimate the strength of the association. I will assess association of age at first pregnancy with ethnicity and endometriosis and if so adjust the estimate accordingly. Furthermore I will adjust for pre-stated confounders (diet, medication). I will assess the magnitude of residual confounding. I will then estimate the strength of the association when adjusted for these confounders.
- studying association between ethnicity, age of first pregnancy and development of) endometriosis
- making a contribution to filling the gap in knowledge about risk factors and effect modifiers for endometriosis
- information may help women to make informed decision about (timing of) their pregnancy
- contributing to scientific, publicly available body of knowledge on endometriosis
- personal learning objectives; final master thesis
Study coordinator London School of Hygiene and Tropical Medicine