New Data Add to Puzzle About Antidepressant, Youth-Suicide Link

A new report suggests that child and adolescent Medicaid patients who have been treated for major depression in an inpatient setting and are prescribed antidepressants may be at a significantly increased risk of a serious suicide attempt or death in the 60 days following discharge, compared with similar patients who are not prescribed antidepressant medications.

The relevance of the study is, however, open to interpretation in part due to limitations of the study design, which did not rule out the possibility that those taking antidepressant medication were more severely ill than those not receiving medication.

“These findings,” said the report's lead author, Mark Olfson, M.D., M.P.H., a professor of psychiatry at Columbia University College of Physicians and Surgeons and a researcher at the New York State Psychiatric Institute, “serve to reinforce some of the concerns that have been raised [regarding an association between antidepressants and suicidality] and really emphasize the importance of paying close attention to changes in mood and actions of severely depressed young people as they start antidepressant medication.”

For several years now, researchers have been striving to answer the question of whether the possible association between antidepressant medications and increased risk of suicidality in children and adolescents extends beyond less-severe expressions of suicidal thoughts or behaviors to serious suicide attempts requiring medical intervention—or ultimately to suicide deaths.

Numerous reports have postulated that the expression of suicidal thoughts and/or behaviors—such as those now addressed by the labeling mandated by the Food and Drug Administration (FDA) accompanying all antidepressant prescriptions—is a different phenomenon than serious suicide attempt and completed suicide. Yet, it is still not clear whether the reported association of the medications with FDA-defined suicidality crosses the hypothesized divide between the two phenomena.

National Case-Control Sample

The study, which appeared in the August Archives of General Psychiatry, analyzed data from the Centers for Medicare and Medicaid Services' (CMS) national Medicaid Analytic Extract Files (MAEF). The MAEF includes comprehensive claims data for Medicaid beneficiaries in all 50 states.

The work was funded by grants from the National Alliance for Research on Schizophrenia and Depression, the Agency for Healthcare Research and Quality, the American Foundation for Suicide Prevention, and a private foundation, the Carmel Hill Fund.

“This was a case-control study of children and adults in Medicaid in which we completed four different analyses. We looked at two outcomes each, in both adults and kids, using the same criteria for both groups,” Olfson told Psychiatric News.

Using the Medicaid claims data, Olfson and his colleagues started with a study population of all patients aged 6 to 64 who had one or more hospitalizations for the treatment of a depressive disorder between January 1, 1999, and December 31, 2000. They then narrowed the sample by excluding patients who were pregnant, had bipolar disorder, schizophrenia, other psychoses, mental retardation, or dementia/delirium. Finally, they refined the cohort further to those beneficiaries who filled at least one prescription for any medication during the two-year study period.

Suicide Attempts Defined

Cases were defined by a diagnostic claim for a suicide attempt that occurred within 60 days following discharge from a hospital for the treatment of major depression. The suicide attempt was further defined as a self-inflicted injury or intentional overdose of medication serious enough to result in treatment in either an emergency department or inpatient setting.

A total of 784 cases with suicide attempts were then matched to 3,635 controls. Each case was matched to up to five control subjects based on date of discharge from an inpatient stay for the treatment of major depression. Matching was also based on age, sex, race/ethnicity, and state providing their Medicaid services.

For suicide deaths, cases were defined as a death within 60 days of discharge from a hospitalization for major depression. Date and cause of death were cross-referenced between the Medicaid database and death certificate data from the National Center for Health Statistics National Death Index. A total of 94 cases with completed suicide were matched, using the same matching criteria as used for suicide attempts, with 435 controls.

Finally, the study population of 4,948 cases and controls was divided into two groups: those aged 18 and younger, and those aged 19 to 64.

Similar but unexpected Findings

Within the defined cases and controls, Olfson and his colleagues—including noted suicide expert David Shaffer, M.D., chief of the Division of Child Psychiatry at Columbia University Medical Center, New York State Psychiatric Institute, and New York Presbyterian Hospital—first looked for any association between suicide attempt and whether the patient had been prescribed and dispensed an antidepressant medication. Then they repeated the analysis for those who died by suicide.

In the adult group, Olfson's team found no significant association between the medications and suicide attempts or suicide deaths. However, in the child and adolescent group, antidepressants were significantly associated with both suicide attempts and completed suicide.

Olfson and his colleagues found that patients in their study population who filled prescriptions for antidepressant medications following an inpatient stay for major depression were 1.52 times more likely to make a serious attempt at suicide (263 cases matched with 1,241 controls) than similar patients who were not prescribed antidepressants.

Patients who filled antidepressant prescriptions following an inpatient stay were 15.62 times more likely to die by suicide (8 cases matched with 39 controls) than similar patients not prescribed antidepressant medications.

While the study's findings are similar to those found by other researchers, Olfson said, he expected to find no real association between the medications and suicide attempts or deaths in either children or adults.

Finding Meaning Is Difficult

Olfson is quick to note that case-control, observational studies have significant limitations, often making it difficult to determine what the findings really mean.

“There's always the nagging doubt with studies like this,” he said, “that even within this defined population of patients following hospitalization for depressive disorder, those receiving antidepressant medications may well have simply been more severely ill.”

Olfson and his team did not have any data indicating the severity of individual patients' symptoms, he noted. Severity was inferred by the necessity for inpatient treatment for major depression. By limiting the analysis to events within the first 60 days following discharge, Olfson added,“ in theory, all of the patients were in the same period of their recovery from that index hospitalization.”

David Fassler, M.D., an APA trustee and child and adolescent psychiatrist, said the study “should be interpreted with caution. The methodology doesn't allow us to draw definitive conclusions in either direction from the results.”

There are many intervening variables, Fassler told Psychiatric News, that could explain the results, including the possibility that patients with more severe symptoms were more likely to receive antidepressant medications.

While numerous studies have found associations between antidepressants and suicidality, Olfson noted there is no clear indication of what factors might help clinicians predict which patients prescribed the medications might be at increased risk versus which patients aren't at risk.

“There is a feeling among clinicians that a personal history or family history of suicide attempt, or bipolar disorder, can increase risk of future suicidal thoughts, behaviors, and actions,” Olfson said. Also helpful, and possibly predictive, is the patient's or family members' response to antidepressant medications, especially in the first few weeks and months of treatment.

“But, bottom line, I don't think at this point the research has done much to give us clear a priori risk factors,” Olfson concluded.“ Clearly, more research needs to be done to establish these or other indicators as true risk factors.”

“Antidepressant Drug Therapy and Suicide in Severely Depressed Children and Adults” is posted at<http://archpsyc.ama-assn.org/cgi/content/abstract/63/8/865>.▪