Post by Former NIMH Director Thomas Insel: Viewing the STARRS Data
Last week, two important research events unfolded without fanfare and without headlines. June 30 marked the end of the first phase of Army STARRS, the largest study of mental health risk and resilience ever conducted among military personnel. July 1 marked the release of Army STARRS data for use by the broad scientific community.
In March of 2014, I blogged about early findings from Army STARRS. The five-year research project received $50 million in funding from the U.S. Army and an additional $15 million from NIMH. Over the course of the study, Army STARRS research teams covered the globe collecting information from tens of thousands of uniformed men and women in an effort to learn more about risk and resilience factors related to suicide.
Although Army STARRS ended on June 30, its mission will continue through Department of Defense funding as the STARRS Longitudinal Study (STARRS-LS), which could turn this project into something like the Framingham Heart Study, except in this case with a focus on mental health outcomes. A second, and potentially equally significant opportunity is the newly released data.
Army STARRS’ data cover an unusually large sample and represent all phases of military life. The Army STARRS research team devoted years to collecting the data and presenting findings to Army leadership. Their work helped dispel many myths about the predictors of suicide and led to the development of statistical methods for identifying small groups of soldiers with very high predicted risk for suicide and/or other adverse outcomes. As just one example, the study was able to identify the 5% of highest-risk service members who accounted for 40% of Army suicide deaths. Making these data available to additional researchers will lead to new analyses and hopefully even more discoveries.
The Army STARRS data released on July 1, 2015, come from more than 60,000 soldiers who participated in one of two study components: the New Soldier Study (NSS), which collected data from soldiers during their first week in the Army, and the All Army Study (AAS), which collected data from established soldiers representing all facets of Army life in the U.S. and overseas. All of the data are free from identifiers so the confidentiality of the participants will not be compromised.
The data are available through a restricted access web portal (not for download). Interested researchers can apply to the Inter-university Consortium for Political and Social Research (ICPSR). Copies of the NSS and AAS questionnaires and a link to the ICPSR application are available on the Army STARRS website.
The openly accessible ICPSR catalogue site provides a range of metadata tools to help interested researchers decide whether to proceed with the application for restricted access to AAS and NSS survey data. If, after reviewing these metadata, a user decides to apply for restricted Virtual Data Enclave (VDE) access, there is a link to initiate the application process.
NIMH and NIH believe final data from NIH-supported research like Army STARRS is a “public good” and that sharing research data with the wider scientific community is essential for expediting the translation of research results into knowledge, products, and procedures to improve human health.
Army STARRS data are only one of several recent new opportunities for learning from NIMH funded studies. The Psychiatric Genomics Consortium has data from over 170,000 volunteers, the National Database for Autism Research has data from over 85,000 volunteers, and the Human Connectome Project has released data from 500 subjects (with 700 more expected later this year).