Testing Teens’ True Feelings About Suicide

One of the most vexing problems in attempting to understand and treat suicide-prone adolescents is that one of the times they are most likely to succeed in taking their own lives is immediately after they’ve been discharged from the hospital. In other words, right after they’ve assured everyone they’re just fine.

This is a problem Matthew K. Nock, a Center-affiliated faculty member, would like to solve—sooner than later. Suicide is the third leading cause of death among U.S. adolescents, according to the federal Centers for Disease Control and Prevention, and rates of suicidal behavior are virtually unchanged over the years, as Nock’s research group has shown.

Finding a solution, however, is not simply a matter of encouraging the teenagers to be more forthcoming about their true feelings, says Nock, who is Director of the Laboratory for Clinical and Developmental Research and Professor in the Department of Psychology in Harvard’s Faculty of Arts and Sciences (FAS). They may not, in fact, feel suicidal at discharge time. “Suicidal thoughts are transient in nature,” says Nock. “They often appear for minutes, hours, or days, and then go away, only to return at some other unexpected time.”

The disconnect between what the adolescent patients say and what they do may also be a function of wanting to avoid being hospitalized or not having a full awareness of the thoughts that drive their behavior. Consequently, relying on verbal self-reports—a traditional method of assessing suicide risk—is very problematic, Nock explains.

nock-pullquote.jpgA better way to measure suicidal thoughts, then, would not rely on what an adolescent says about himself or herself. In the past, researchers have looked for biological markers or risk factors for suicide, Nock says, but, he notes, “the pathway to suicide is such a complex psychological process that one biological factor or genetic component is not going to specifically predict it.”

With funding from the Center, Nock is working to develop ways of using adolescents’ behavior to ferret out whether they have suicidal feelings.

One test he is using to do so is the Implicit Association Test (IAT). Originally developed by psychologist Anthony Greenwald of the University of Washington, the IAT has been used extensively by Mahzarin Banaji, who is the Richard Clarke Cabot Professor of Social Ethics in the psychology department of FAS. Social psychologists, Nock says, have used the test to determine whether people have hidden biases about such things as race, ethnicity, gender, or obesity. (See Project Implicit,

Nock has collaborated with Banaji to adapt the IAT for use on adolescents with suicidal thoughts. His version, the Suicide Implicit Association Test (S-IAT), is given on a computer, takes about five minutes, and uses an individual’s reaction time to suicide-related words and images to examine automatic mental associations about suicide.

In the test, the computer program presents words related to suicide (for example, “death,” “funeral,” “self-harm”) and words related to life (for example, “live,” “survive,” “health,”) as well as words related to concepts of “me” and “not me.” The premise of the test is that people are faster at connecting words their minds unconsciously associate as similar, as compared with words they consider dissimilar.

For Nock, adapting the test “seemed to me to be the perfect marriage between a longstanding clinical problem and a new scientific method of measuring implicit thoughts.”

The Center is funding a newer version of the S-IAT that is being administered to suicidal adolescents at Children’s Hospital Boston. Adolescents are given the test when they are admitted to the psychiatric inpatient unit and again when they leave. They are also contacted by phone three months later. The idea behind the current study, Nock says, is to look at three things: 1) how well this adaptation of the IAT can distinguish between those who attempt suicide and those who do not, 2) whether a teenager’s suicidal thoughts change over the course of a hospital stay, and 3) whether the test can better predict who attempts suicide after leaving the hospital.

If the test can provide a reliable way to measure suicidal thoughts, Nock says, “it would be a useful clinical tool to bridge the science-practice gap, to use some recent advances in science and translate that in a way that is directly clinically useful.”

The need to close that gap is well-known to Nock, who is part of a three-person Center working group known as the Child Mental Health Network. The trio is working to bring together the knowledge that has been gathered through years of mental-health research and the practices used with children and adolescents in clinical settings. As the group wrote in its 2008 concept paper, “Those who deliver mental health services and those who study them have operated independently, with little interaction and little impact on one another.” As a practical matter, that means that too many kids don’t get the treatment that researchers know could help them.

By bringing together scientists, practitioners and policymakers, the network hopes to generate, integrate, communicate, and apply the science of children’s mental health to influence policy and practice and inform public understanding by making scientific advances more transparent.

Nock, who is also a member of the Center’s steering committee, says he appreciates the Center’s focus on translating science into better policies and practices: “It provides a way for the scientific work that we’re doing to have a very real impact on the world.”

-- Millicent Lawton

Photo by Fred Field

View full series of Faculty Spotlights  >>