Data Dictionary
A comprehensive cognitive and clinical data dictionary in standard LONI format is available under Study Docs > Study Dictionary and Documents.

Pairing Cognitive & Clinical Data with Images
The study structure of WRAP is unique. Imaging and other biomarkers have generally been obtained for WRAP participants as part of separately funded substudies. As such, there is not a single "Visit Number" that allows for appropriate pairing of cognitive & clinical visits with imaging measures. Because of this, we provide information on the relative timing of all events to allow investigators to create the pairings that are most appropriate for their research question.

• In the Study Data tables, the table containing the primary keys for joining cognitive & clinical tables together is called fqryStatisticalData. Most other tables can be joined with this one either on wrapnum (for subject-level information) or on both wrapnum and VisNo (for subject-visit level information).
• In fqryStatisticalData, you will also find a field called Days_Since_Enrollment. This establishes the timescale for joining the cognitive & clinical data with events that did not happen at a main study visit, like imaging. For VisNo=1, Days_Since_Enrollment should be 0.
• In the table called Imaging_metadata, you will find information about the images, including their timing. This table also includes a Days_Since_Enrollment field, which is on the same time scale as the same-named field in fqryStatisticalData. It does not include a VisNo field, as imaging visits and cognitive/clinical visits are not the same. There is, however, an imaging_visit field, which for participants with more than one image will index their relative position in the series. Importantly, imaging_visit=1 will not generally have a value of 0 in this field. Frequently the first image will have been acquired many years after the initial study visit.
• Investigators will not usually want to perform a strict join between images and cognitive/clinical visits using wrapnum and Days_Since_Enrollment, as very few visits happen on the same day. Instead, it is better to think about an approach to joining the data that best supports your research question. Here are some examples of approaches that some in our group have taken:
  1. Select the first available image and pair with the full range of cognitive visits. Useful if your question is, "How do the earliest available brain biomarkers relate to cognition?"
  2. Select the first available image and pair only with the cognitive visits that happened at or after the image. Useful if your question is, "How does a biomarker at time t0 relate to subsequent cognition?"
  3. Select the last available image and pair with the full range of cognitive visits. Useful if your question is, "How does the best/most current information about biomarker phenotype relate to retrospective cognition?
  4. Select longitudinal pairs of images and cognitive/clinical visits, subject to some time constraint (e.g. visits must happen within 1 year of each other). If you select this option, you may want to consider what to do when a single visit of one type is the closest match to more than one visit of the other type.