How to ensure inclusivity in large-scale general population cohort studies? Lessons learned with regards to including and assessing sex, gender and sexual orientation.

Despite recent advances in awareness and measuring of sex, gender and sexual orientation in large-scale cohort studies, the three concepts are still gaining relatively little attention, may be mistakenly equated, or non-informatively operationalized. The resulting imprecise or lacking information hereon in studies is problematic, as sex, gender and sexual orientation are important health-related factors. Omission of these concepts from general population cohort studies might dismiss participants’ identity and experiences, and pushes research on sexual or gender minority populations towards purposive sampling, potentially introducing selection bias. It also reinforces the unintentional notion of irrelevance of these concepts to health research, ultimately disadvantaging sexual and gender minority populations. Similarly, a lack of uniform measures on sex, gender and sexual orientation hampers multi-cohort studies in which data from multiple studies are combined, facilitating increased statistical power. This paper discusses the encountered pitfalls and lessons learned on including and assessing sex, gender and sexual orientation in large-scale general population cohort studies, exemplified by the Dutch Lifelines Cohort Study. Additionally, we propose hands-on strategies on how to operationalize these concepts in an inclusive manner that is useful for large-scale general population cohort studies.