Acceptability and Feasibility of Using Hair Samples for Chronic Stress Measurement Among Transgender Women in Brazil.

Introduction: The ability to objectively measure chronic stress has important implications for research, prevention, and treatment. Cortisol is currently the most used biological marker in the investigation of stress and can be measured via blood, saliva, and urine; however, these methods have disadvantages. The measurement of cortisol in hair is a more recently developed method that quantifies the cumulative production of cortisol over longer periods of time. Given the potential benefits of hair as a chronic stress biometric, research with this novel method is burgeoning, yet rarely involves transgender ("trans") populations, despite high levels of reported stress among trans people due to experiences of stigma and discrimination. Since hair is a key part of gender presentation, trans people might be less likely than cisgender people to donate hair for research. To explore the feasibility and acceptability of hair collection for use as a stress biometric with trans women, we nested a study into an ongoing clinical trial in São Paulo, Brazil, "Manas por Manas" (Sisters for Sisters). Here, we describe the hair biometric substudy protocol, as well as the feasibility and acceptability of collecting hair in the study cohort.

Methods: We randomly selected a subsample (n = 180) from the Manas por Manas cohort (n = 392), all of whom are trans women, age 18 or older. We messaged participants via phone, WhatsApp, or social media for at least three attempted contacts. Study visits included four components: (1) video introduction, including a demonstration of hair sampling; (2) informed consent; (3) a brief survey with the validated Short Stress Overload Scale (translated to Portuguese) and questions on hair care that could moderate stress hormone results; and (4) hair sample collection. Hair samples were collected and stored using validated protocols. Participants were reimbursed for travel costs.

Results: Between April and December 2022, we messaged with 143 individuals out of the 180 sampled (79%) and invited them to participate in the study. Of those invited to participate, we scheduled study visits with 102 people (71.3% of those invited to participate), of whom 100 attended their study visits and completed all activities. Two people did not attend their study visits and stopped communication. Of those who were invited to participate and declined a study visit, four individuals declined due to the hair sample collection procedures (2.8% of those invited to participate). Other reasons for declining to participate included having moved (n = 7), lack of time (n = 11), not interested in research participation (n = 8), or unknown/stopped responding to messages (n = 11). Most participants reported that they chemically treated their hair to bleach, color, or straighten it, which could impact laboratory assays.

Conclusion: We found hair sampling for stress measurement to be feasible and acceptable to our participants. We successfully completed all study activities for our desired sample size, and most recruited individuals volunteered to participate. Reasons provided for declining study participation reflected general barriers to research participation, with only four people declining due to hair sample collection procedures.