Skip to main content

Transforming the understanding
and treatment of mental illnesses.

The Making of Army STARRS: An Overview

Behavioral health and suicide are issues that affect Americans from all walks of life. The Army Study to Assess Risk and Resilience in Servicemembers (Army STARRS) is an unprecedented research study dealing with a topic critical to the Army and to the country—suicide prevention.

Historically the suicide rate among Army personnel has been below that of the civilian population. Since 2002, however, the suicide rate among soldiers has risen significantly, reaching record levels and prompting the Army to form a partnership with the National Institute of Mental Health (NIMH) to seek independent academic scientists who could design and implement a sufficiently large and complex research study to address the issue. The NIMH awarded a grant to a research team of investigators from the Uniformed Services University of the Health Sciences, Harvard Medical School, the University of Michigan, and the University of California, San Diego to design and conduct Army STARRS.

Army STARRS is the largest study of mental health risk and resilience ever conducted among military personnel and, following a one-year no-cost extension, will run through June 2015. Each of the components of the study is confidential and strictly voluntary. Since the study began in 2009, Army STARRS has worked at more than 75 locations in the United States and around the globe. The research team has collected data from more than 100,000 soldier volunteers in all phases of Army service including new soldiers in their first week in the Army, established soldiers, soldiers pre- and post-deployment (see below), soldiers in theater, and soldiers who have been hospitalized for self-harm (see below). The Army STARRS data-collection phase ended in April 2014 and the research team has now turned their focus to data analyses.

In an effort to find risk and protective factors that affect soldiers’ well-being, the Army STARRS team has undertaken some enormous tasks. For example, they created a secure database with more than one billion Army and Department of Defense (DoD) data records and 3,000 types of information. The team is analyzing these administrative data as well as new information collected directly from soldier volunteers. All of this has been done, and will continue to be done, with extreme care for the security and confidentiality of each soldier who volunteered to participate.


Every research study faces challenges and opportunities to develop new methods. This is also true with Army STARRS. The Making of Army STARRS series highlights some of the challenges and opportunities that have been part of this project.

Coordinating Cultures

The academic and military worlds may speak different languages and approach a situation from different perspectives but when the goal is the same, everything comes together. The Army STARRS research team has broadened its understanding of military life and acronyms just as the Army’s team has added to its knowledge of statistical sampling methods and research lingo. The commitment to making Army STARRS a success is what keeps everyone moving forward.

Studying a Rare Event

Suicide is a very rare and complicated event. In fact, on average, less than 1 in 5,000 people die by suicide. In addition, there are few things, if any, common to all suicides. For example, although some risk factors such as depression or failed relationships often precede suicide, most soldiers who experience these things never try to take their own lives. Because suicide is so rare, researchers must look at very large groups of people to draw valid scientific conclusions. Army STARRS has more than 100,000 participants.


Confidentiality is an extremely important element of this project. In every part of the study, in every phase, information that could identify an individual soldier was removed from survey responses, administrative data, historical data, and biological specimens. Survey ID numbers or scrambled identifiers were attached to different types of data instead of names, addresses, or Social Security Numbers.

Connecting the Dots

Army STARRS collected information directly from tens of thousands of soldiers. The team uses a computer program to link soldiers’ Army STARRS data to de-identified administrative data from Army and DoD sources. This is done with soldiers’ permission and in compliance with all legal and privacy protections. The computer program makes it possible to link survey and administrative records in such a way that the participating soldiers cannot be personally identified through their responses in the research data. Because of this linked data, Army STARRS will support more detailed studies of both risk and protective factors for suicide than has been possible in the past.

The Big Picture

To get a complete picture of soldiers and their experiences, Army STARRS has five primary parts. The components look at both historical and current data, and several are represented in The Making of Army STARRS series. Findings from each of the five components may help researchers and the Army identify opportunities to improve soldier well-being and reduce the risk of self-harm.

The five primary components are:

  • Historical Administrative Data Study: Using data from ~40 data sources from the Army and DoD, this component examines administrative records from more than 1.6 million soldiers who served on active duty at any point from 2004-2009. More than one billion records have been stripped of information that might identify an individual soldier. Investigators are using the data to help identify periods in a military career when soldiers might be at particularly elevated-risk for behavioral health challenges such as during deployment to a war zone and the periods immediately before and after that deployment.
  • All Army Study (AAS): This Army STARRS component assesses data from thousands of active duty soldier volunteers (including activated Reserve Component and Army National Guard soldiers). The AAS is looking at psychological and physical health; events encountered during training, combat, and non-combat operations; and life and work experiences across all phases of Army service. Researchers use the information to determine how various factors affect soldiers’ psychological resilience, mental health, and risk for self-harm (e.g., suicide).
  • New Soldier Study (NSS): Army STARRS researchers invited new soldiers at the start of their basic training program to volunteer for the New Soldier Study. Information on new soldiers’ health, personal characteristics, and prior experiences as they begin their Army service will help the research team understand the impact of factors that exist prior to entering the Army.
  • Soldier Health Outcomes Study (SHOS-A & B): This component is actually two sub-studies. Each study compares participants who have exhibited suicidal behavior with those who have not. SHOS-A focuses on soldiers who attempted suicide and were admitted to a medical treatment facility. SHOS-B focuses on soldiers who died by suicide and involves interviews with next of kin and Army supervisors. Both studies will attempt to identify characteristics, events, experiences, and exposures that predict negative (or positive) health and behavior outcomes. More information about SHOS-B is available here.
  • Pre/Post Deployment Study (PPDS): The PPDS component focuses on the effects of deployment to a theater of combat operations. The research team is examining the relationship between physical and biological changes and suicidal thoughts and actions or related behavioral health issues. The pre-deployment phase concluded in 2012 with more than 9,000 volunteers. These same volunteers were asked if they would be willing to participate in the three-part post-deployment phase, which ran through April 2014.