Intersectional health disparities: the relationships between sex, race/ethnicity, and sexual orientation and depressive symptoms

ABSTRACT

This study uses Intersectionality-informed methods to quantitatively examine experiences of discrimination and depressive symptoms among historically marginalised populations. Using mediation models and data from Project STRIDE collected in 2005–2006, discrimination (attributed to gender and/or sexual orientation and/or race/ethnicity) was modelled as a mediator between identity and depressive symptoms among seven diverse identity-based subgroups (75% sexual minorities and 50% Black or Hispanic). Discrimination partially mediated the effect of identity on depressive symptoms for sexual minority Black women, but only when accounting for discrimination on the basis of all three marginalised identities (woman, Black, and lesbian/bisexual). Sexual minority Black and white men experienced significantly less/less frequent depressive symptoms relative to straight white men, holding constant discrimination. While quantitatively modelling intersectionality is inherently nuanced, this study may serve as a framework for carrying out future quantitative intersectionality-based studies. Ultimately, intersectionality research contributes to the potential for a healthier and more equitable society for all.

KEYWORDS:

• Intersectionality
• depression
• discrimination
• minority stress

Disclosure statement

The authors have no conflicting interests to disclose.

Data Availability Statement

The data are available here: https://www.icpsr.umich.edu/icpsrweb/RCMD/studies/35525/datadocumentation

Notes

1. Project STRIDE data were collected before the issuance of the 2019 APA standards, which advise that studies survey participants’ gender identity as cisgender or transgender and man, woman, or non-conforming/non-binary, while sex refers to the sex that people were assigned at birth.

2. Project STRIDE stratified race/ethnicity by these three races/ethnicities only, thus we were unable to include other races/ethnicities (e.g. Asian American, Indigenous, etc.)

3. Following Valeri and VanderWeele (2013) guidelines handling exposure-mediator interactions, two sets of regression models were fit. The first regressed the mediator(s) on the specified control variables and the identity indicator. This yielded the a path(s) of the mediation model, that is, the effect of the subgroup on respective discrimination. The second regressed the outcome on the specified control variables and the mediator(s), the identity indicator, and the interaction between the mediator(s) and the identity indicator.

4. We tested all possible combinations of ‘summed natural indirect effects’ including the summed natural indirect effect of sexual orientation and race/ethnicity; sexual orientation and sex/gender; and sex/gender and race/ethnicity. Only the sum of the indirect effects that included all three mediators was significant.

Additional information

Notes on contributors

Abby K. Johnson Holm

Abby K. Johnson Holm, M.S., PhD is a student in Applied Social and Health Psychology at Colorado State University. She is employed full time as a data scientist for Hopkins Public Schools where she studies educational disparities.

Ashlie N. Johnson

Ashlie N. Johnson, M.S. is a fifth-year PhD student in Applied Social and Health Psychology at Colorado State University studying eating behaviours and health disparities.

Raeven Clockston

Raeven Clockston is a second year Master of Public Health student at Colorado State University studying epidemiology and health disparities.

Katrina Oselinsky

Katrina Oselinsky is a third year PhD student in Applied Social and Health Psychology at Colorado State University. She studies physical activity and health disparities.

Pamela J. Lundeberg

Pamela J. Lundeberg, M.S., PhD is a psychology professor at Aims Community college where she studies physical activity and health disparities.

Katelyn Rand

Katelyn Rand is a fourth-year undergraduate psychology student at Colorado State University.

Daniel J. Graham

Daniel J. Graham, M.A., PhD is an associate professor at Colorado State University. He studies social-ecological determinants of health and multilevel interventions promoting health behaviour.