3. Mapping the Gender Gradient

​​​

Project leader: Rachel Kimerling, National Center for PTSD, VA Palo Alto Healthcare System

US Department of Veterans Affairs, USA

​

Project group: Rachel Kimerling, National Center for PTSD, USA; Rachel Ward, University of New Mexico Hospitals, University of New Mexico Health Sciences Center, USA; Mckenzie Lockett, National Center for PTSD, USA; Chris Hoeboer, Amsterdam UMC, Netherlands; Miranda Olff, Amsterdam UMC, Netherlands; Soraya Seedat, Stellenbosch University, South Africa; Lena  Oestreich, The University of Queensland, Australia; Sarah Halligan, University of Bath, United Kingdom; Stephanie, Freie Universität Berlin, Germany

​

Collaborate

Collaborators are welcome (CAW)! Please contact Rachel Kimerling, if you are interested to join this project.

​

Background

The prevalence of posttraumatic stress disorder (PTSD) among women is over twice that of men. This disparity is observed across a wide range of populations and nationalities despite a relatively similar phenotypic expression of PTSD. A key challenge in elucidating the role of gender in PTSD is that sex and gender are not mechanisms that drive these differences but rather artificial dichotomies that function merely as group labels. Even when sex assigned at birth is differentiated from gender identity, or when categories are created for genderfluid, transgender, or other nonbinary identities, group comparisons still obscure important information about within-group variability or the mechanisms by which gender impacts health. These group labels are proxies for complex, multilevel, interdependent biological and social factors. The challenge is to develop new models for research designs and approaches that move the binary to model the effects of sex and gender as multidetermined continua.

Gender is widely accepted as socially constructed—by the behaviors individuals perform, the expectations and social reinforcement that shape those behaviors, and the opportunities defined by social structures. All of these influences have the potential to impact exposure to and recovery from traumatic stress. Broader efforts to quantify the health impacts of gender have led to calls for the inclusion of gender as a sociocultural variable (GASV) in a manner similar to policies encouraging research to address sex as a biological variable (SABV). We broadly use the GOING-FWD Framework to organize gender indicators.

Aims

This project explores novel approaches to examining gender and PTSD, moving beyond the binary group labels of male and female to explore the summative impact of gender-linked sociocultural factors. We will explore the utility of machine learning to model the multiple intersecting indicators of gender from theoretical and empirically selected predictors reflecting the roles, relationships, and institutional facets of GASV. Aims are to:

1. Identify theory-based sociocultural indicators of gender.

2. Model categorial gender via statistical models that allow for complex interrelationships among predictors. Models will be used to obtain predicted probabilities that comprise the gender scores.

3. Use gender scores to predict PTSD prevalence or incidence as compared to gender categories.

Method

Collaborators will identify sufficiently large and representative datasets representing their country or region that include both PTSD symptoms or diagnoses and potential indicators of GASV across the breadth of the construct. Though each study will stand alone, we will compare results based on country, gender inputs, and model results to make cross-national inferences about GASV as it relates to the prevalence of PTSD.

Predictors: We suggest GASV predictors that do not directly represent trauma exposures, biological factors (e.g. ever having been pregnant), or other intrinsic indicators to avoid indicators that are so collinear with gender categories that they may obscure social indicators of roles, relationships, and institutional aspects of gender. We recommend considering variables, or creating variables such as:

· Direct indicators: examples include social roles, such as caregiving, income, education

· Ratios or other indicators of gender-based disparities: for example, ratio of personal income, relative to the mean of opposite sex peers within the same occupational class

· Composite measures: such as social networks, minority stress, role strain, sex roles

· Moderators such as race, ethnicity, age or other factors potentially define intersectional effects

Status: Starting up

​​

​

​

​

​

​

​​