Representativeness of the Danish Blood Donor Study relative to the general population: a cross-sectional assessment
摘要
Cohorts that are non-representative of their target population pose a critical challenge, limiting generalizability and potentially yielding misleading effect estimates. To this end, we conducted a cross-sectional assessment of the representativeness of the Danish Blood Donor Study (DBDS)—a nationwide cohort of 170,000 individuals recruited during visits to blood donation centers across Denmark—relative to the Danish general population aged 18–74 years. Using national register data, we compared demographic, socioeconomic, and health-related characteristics of DBDS participants to those of non-DBDS individuals as of December 31, 2021. Our findings revealed that DBDS participants were more likely to be well-educated, live with an adult of the opposite sex, and have two children. Although individuals living alone were generally underrepresented, young males (18–24 years) and females (18–44 years) in this group were notably overrepresented. In addition, those with high incomes or with employment in finance or insurance were strongly overrepresented, whereas immigrants and rural residents were underrepresented. Participants were less likely to have any recently recorded diagnoses or redeemed prescriptions across most major disease and treatment categories. In particular, individuals with mental, behavioral, or hematological disorders were markedly underrepresented, whereas females (25–44 years) with diagnoses related to pregnancy, childbirth, or the perinatal period—or with redeemed prescriptions for urogenital or hormonal medications—were overrepresented. This characterization of DBDS participants provides context for interpreting results obtained from the cohort and underscores the need for methodological approaches to improve the generalizability of findings derived from non-representative samples.