Brain-wide association studies using magnetic resonance imaging have come under scrutiny for producing results that are often difficult to replicate. However, a recent study published in Nature has shed light on how careful study design can significantly enhance the reliability of this type of research.
The study, led by Kaidi Kang, a biostatistics Ph.D. student, and Simon Vandekar, Ph.D., an associate professor of Biostatistics, analyzed data from over 77,000 brain scans across 63 different studies. The researchers discovered that by strategically selecting study participants to encompass a wider range of characteristics being measured, studies can yield more dependable results. For instance, when examining age-related brain changes, including participants from both younger and older age groups can lead to more reliable findings compared to random sampling or focusing solely on middle-aged individuals.
Moreover, the study highlighted that collecting multiple brain scans from the same individual over time can enhance reliability for certain brain measurements while possibly reducing it for others, depending on the specific focus of the study. These insights provide valuable guidance for scientists designing future brain imaging studies, enabling them to achieve reliable results with fewer participants.
In addition to Kang and Vandekar, researchers from six institutions in the U.S. and the United Kingdom contributed to the study, including Jiangmei Xiong, Megan Jones, Ran Tao, and Jonathan Schildcrout from Vanderbilt University.
For further information, the study titled “Study design features increase replicability in brain-wide association studies” can be accessed in Nature (DOI: 10.1038/s41586-024-08260-9).
This groundbreaking research not only enhances our understanding of brain structure and function but also provides valuable insights for improving the reliability of future brain imaging studies, ultimately advancing our knowledge of human behavior and health.
