Cohort study is one of the important epidemiological methods which plays an irreplaceable status and role in etiological study.Using cohort study design,we can accurately and continuously collect genetic and environmental information,and identify and validate omics biomarkers to provide evidences for precision public health and medicine.However,results from a new cohort would not be available for at least ten years,as five years would be needed for funding,planning and enrolment,and another five for following up even the earliest analyses of the most common diseases;results for most cancers would take longer,with an unaffordable budget for many research investigators or institutions.That brings an alternative strategy of using existing cohort studies by sharing data between each other.Data sharing of cohort studies would be beneficial in many ways.Data sharing of cohort studies has the potential to make large samples unattainable in a single study,increase statistical power,enable more accurate and detailed

作者：杨羽；赵厚宇；詹思延

来源：北京大学学报（医学版） 2018 年 50卷 2期