Participant selection bias in women’s health studies may mask earlier onset menopause for Black and Hispanic women, according to new research led by University of Michigan.
Researchers say that failure to account for weathering led to exclusion of many Black and Hispanic women from the Study of Women’s Health Across the Nation (SWAN) cohort and misses critical racial differences in menopausal age.
Established in 1994, SWAN examines midlife health and menopausal transition to improve interventions and educational programs for women of this life stage. Weathering, a framework developed by U-M School of Public Health professor Arline Geronimus, posits earlier health deterioration of oppressed and exploited populations due to social influences and chronic stress. The research is published in the International Journal of Epidemiology.
"We were able to quantify the racial differences in the rate of exclusion from SWAN due to earlier menopause, and then statistically account for it in SWAN’s data,” said Alexis Reeves, a postdoctoral scholar at Stanford University’s School of Medicine who conducted the work while a doctoral student at the U-M School of Public Health. "We found that Black and Hispanic women had statistically significant earlier natural, and particularly surgical, menopause than white women. The study suggests that this common bias may lead to underestimation of racial disparities in health and aging, and is important to consider in further research."
According to the findings, Black and Hispanic women reached menopausal age approximately 1.2 years earlier than their white counterparts when exclusion due to weathering is accounted for, whereas SWAN’s original data found little to no racial differences in menopausal age. This study is one of the first to consider that weathering biases an understanding of racial disparities, particularly with data to support it.
"Mistrust of the medical system due to historical injustices is an important part of the lack of inclusion of minoritized populations in studies. However, this study suggests that eligibility criteria—and the given age for inclusion into studies—set by researchers themselves also plays an important part in exclusion of minorities from studies," Reeves said.
"The implications of these findings are incredibly important to understanding the true burden of racial disparities in women’s health and indicate that researchers need to be more thoughtful about eligibility criteria and the potential for underestimating racial disparities in longitudinal health studies," said Siobán Harlow, professor emerita of epidemiology at the School of Public Health and senior author of the study.
The researchers are hopeful that bringing attention to mitigating selection bias can help address these disparities. "Accounting for data biases in future studies will allow us to better understand and address the negative health outcomes of these marginalized populations," Harlow said.
Additional U-M authors of the study include Michael Elliott and Carrie Karvonen-Gutierrez.
Journal
International Journal of Epidemiology