Does psychiatric treatment prevent suicide?

Over the past 2 decades, more and more people have been treated with antidepressant medications. In the same period of time, suicide rates have gone up – not down. To those of us who treat patients, this fact is both surprising and perplexing. It seems that suicidal thoughts are a common feature of major depression, and when the depressive symptoms abate with treatment, the suicidal thoughts dissipate. Intuitively, it seems that treating depression on a larger scale should prevent suicides, but we still don’t know that conclusively.

According to the National Health and Nutrition Examination Survey (NHANES), 11% percent of Americans over the age of 12 are taking an antidepressant medication. In women aged 40-59, this number is 23%. Of those taking antidepressants, only one-third have seen a mental health professional in the past 12 months. What also is striking is that for people surveyed with symptoms of severe depression, only one-third were on medication.

In 2013, just over 40,000 Americans died of suicide. From 2000 to 2013, the suicide rate per 100,000 Americans has steadily increased from 10.4 to 12.6 per 100,000 people. While we know that people with psychiatric illnesses have higher rates of suicide compared with the general population, what we don’t know is whether the people dying are the same people who are getting treatment.

Thinking about this gets very difficult. It has been estimated that 90% of those who die of suicide have suffered from a mental illness. This figure includes those who were treated, untreated, and previously treated, but the studies have methodologic inconsistencies and that 90% estimate may not be accurate. Certainly, however, people die of suicide for reasons that have nothing to do with psychiatric illness, and we do know that impulsive responses to distressing circumstances are a factor, especially when a lethal method is easily available.

Several studies have shown that antidepressant use, particularly in older adults, may be associated with a decrease in suicidality. The studies often combine suicide attempts and completions. The issues with younger patients are more complicated, and in recent weeks, the reanalysis of the 2001 Paxil 329 study has again raised issues about the safety of certain antidepressants in children and adolescents. The data for all these studies are both confusing and contradictory, and are not easy to examine or interpret.

We also don’t know what role psychotherapy plays. A study done at Johns Hopkins Bloomberg School of Public Health looked at the follow-up for 65,000 people in Denmark who had attempted suicide and found that the rates for completed suicide dropped if the patient received a short course of psychosocial therapy at a suicide prevention center. But again, this study looked at a select group of people who had already attempted suicide.

I began to think it might help to ask these questions in a closed system where patients could be tallied with regard to requests for treatment, what type of treatment was provided, and even access to autopsy results. The U.S. Department of Veterans Affairs seemed to be a source where such answers might be found. It has been reported that 22 veterans a day die by suicide, and many veterans get their care in VA facilities, with VA pharmacy benefits, and treatment effects can, in theory, be studied.

Hoping to get a better sense of the relationship between treatment and suicide, I met with Robert Bossarte, Ph.D., director of the Epidemiology Program in the VA’s Office of Public Health. His career has been focused on suicide prevention.

The first thing Dr. Bossarte did was dissuade me of the idea that the VA is a closed system. Not all veterans receive lifelong benefits from the VA, and the formulas for determining who is entitled to what benefits, and for how long, is rather complicated. Dr. Bossarte also noted that some patients go outside of the system for their care.

“What we do know is that among those who have used VA services in the previous year, about 2,000 veterans a year die from suicide. It’s been hovering around that for the past decade,” Dr. Bossarte explained, again emphasizing that many veterans do not receive care at VA facilities. “We published a report in 2012 where we estimated 22 veterans a day die by suicide and that caught fire, but it is purely an estimate. The truth is we have no idea what the real count is, because until we began working on the Suicide Data Repository there was no national register of veteran mortality and there has been no way for us to know.”

Dr. Bossarte anticipates that the VA will have the data necessary to calculate more accurate statistics by the end of the year. “More than 70% of veteran suicide are in people over age 50; but the rates are going up most among the youngest.”

A notable drop in veteran suicide rates for those who used services occurred between 2001 and 2003, and that decrease remains unexplained; it preceded later changes in mental health services and enhanced suicide prevention programs. Dr. Bossarte also pointed out that just under half of veterans who die from suicide have no mental health diagnosis, despite yearly screening to identify people who may be suffering from posttraumatic stress disorder, alcohol-related disorders, and depression.

“The attention on veteran suicide started around 2007,” Dr. Bossarte explained. “Mark Kaplan published a study using publicly available mortality data; those who reported they were veterans were twice as likely to die of suicide as those who were not.”

While this sparked interest in veteran suicide, it’s important to note that a replication of that study in 2012 by Matthew Miller did not have the same findings.

“Then, in 2008, for the first time in recent history,” Dr. Bossarte continued, “the suicide rate among active duty military personnel exceeded that of the general population. Traditionally, rates of suicide in this population have been 40%-50% lower than in the general population. The increased rate was seen primarily in the Army and Marines. Serious mental illness may make people ineligible for military service, as can violent and disruptive behavior – things that are associated with suicide – so you tended to get a healthier population in the military.”

Dr. Bossarte noted that there was conjecture that increased suicide rates among active military might be related to more waivers that allowed people to enlist who would not ordinarily be eligible, and/or to higher rates of deployment. He went on to talk about Army STARRS (Army Study to Assess Risk and Resilience in Servicemembers).

“STARRS devoted $50 million over 5 years to the largest suicide study and did not find an effect of waivers. They did report a higher suicide rate among those who were deployed, however. But then Tim Bullman in my office looked at suicide rates 7 years after separation from service, and he reported a higher suicide rate among those who were never deployed.” The VA studies, I quickly realized, were also confusing and contradictory.

The VA has greatly expanded its mental health and suicide prevention services. For veterans overall, suicide rates have stabilized, but they have not decreased. For those veterans with psychiatric disorders, however, the suicide rates have gone down.

“When you ask ‘does treatment matter?’ it’s so hard to disentangle psychotherapy from pharmacotherapy. Over the past decade, we’ve seen a significant decrease in the suicide rate among those veterans with mental health disorders. We’ve looked at suicide rates every way you can think of. One thing we do know is that the better the relationship with the clinician, the lower the suicide risk.”

We talked about the role of hospitalization in preventing suicide, and Dr. Bossarte noted that the highest risk for suicide is immediately following hospital discharge.

“We are looking at people hospitalized after their first-ever suicide attempts and rates of mortality, including suicidal behavior, for 1 year after discharge. In very preliminary findings, we didn’t see any difference in the outcome for either all-cause mortality or repeat suicide attempts in those who were hospitalized, compared to those who were not. We don’t yet know about completed suicide.”

I left my discussion with Dr. Bossarte with more questions than answers. We have reason to believe that treatment helps, but we still don’t know which treatments help which people, and we do know that treatment doesn’t prevent suicide in every patient. In a culture where “treatment” has come to be equated with “prescribing” and is often based on a checklist of symptoms done by a primary care clinician, one might wonder if combining psychotherapy and medication – an increasingly rare offering – might have a better outcome. Simply put, for a problem that prematurely takes more than 40,000 lives a year, we know much too little.

Dr. Miller is a coauthor of “Shrink Rap: Three Psychiatrists Explain Their Work” (Baltimore: Johns Hopkins University Press, 2011).

Over the past 2 decades, more and more people have been treated with antidepressant medications. In the same period of time, suicide rates have gone up – not down. To those of us who treat patients, this fact is both surprising and perplexing. It seems that suicidal thoughts are a common feature of major depression, and when the depressive symptoms abate with treatment, the suicidal thoughts dissipate. Intuitively, it seems that treating depression on a larger scale should prevent suicides, but we still don’t know that conclusively.

According to the National Health and Nutrition Examination Survey (NHANES), 11% percent of Americans over the age of 12 are taking an antidepressant medication. In women aged 40-59, this number is 23%. Of those taking antidepressants, only one-third have seen a mental health professional in the past 12 months. What also is striking is that for people surveyed with symptoms of severe depression, only one-third were on medication.

In 2013, just over 40,000 Americans died of suicide. From 2000 to 2013, the suicide rate per 100,000 Americans has steadily increased from 10.4 to 12.6 per 100,000 people. While we know that people with psychiatric illnesses have higher rates of suicide compared with the general population, what we don’t know is whether the people dying are the same people who are getting treatment.

Thinking about this gets very difficult. It has been estimated that 90% of those who die of suicide have suffered from a mental illness. This figure includes those who were treated, untreated, and previously treated, but the studies have methodologic inconsistencies and that 90% estimate may not be accurate. Certainly, however, people die of suicide for reasons that have nothing to do with psychiatric illness, and we do know that impulsive responses to distressing circumstances are a factor, especially when a lethal method is easily available.

Several studies have shown that antidepressant use, particularly in older adults, may be associated with a decrease in suicidality. The studies often combine suicide attempts and completions. The issues with younger patients are more complicated, and in recent weeks, the reanalysis of the 2001 Paxil 329 study has again raised issues about the safety of certain antidepressants in children and adolescents. The data for all these studies are both confusing and contradictory, and are not easy to examine or interpret.

We also don’t know what role psychotherapy plays. A study done at Johns Hopkins Bloomberg School of Public Health looked at the follow-up for 65,000 people in Denmark who had attempted suicide and found that the rates for completed suicide dropped if the patient received a short course of psychosocial therapy at a suicide prevention center. But again, this study looked at a select group of people who had already attempted suicide.

I began to think it might help to ask these questions in a closed system where patients could be tallied with regard to requests for treatment, what type of treatment was provided, and even access to autopsy results. The U.S. Department of Veterans Affairs seemed to be a source where such answers might be found. It has been reported that 22 veterans a day die by suicide, and many veterans get their care in VA facilities, with VA pharmacy benefits, and treatment effects can, in theory, be studied.

Hoping to get a better sense of the relationship between treatment and suicide, I met with Robert Bossarte, Ph.D., director of the Epidemiology Program in the VA’s Office of Public Health. His career has been focused on suicide prevention.

The first thing Dr. Bossarte did was dissuade me of the idea that the VA is a closed system. Not all veterans receive lifelong benefits from the VA, and the formulas for determining who is entitled to what benefits, and for how long, is rather complicated. Dr. Bossarte also noted that some patients go outside of the system for their care.

“What we do know is that among those who have used VA services in the previous year, about 2,000 veterans a year die from suicide. It’s been hovering around that for the past decade,” Dr. Bossarte explained, again emphasizing that many veterans do not receive care at VA facilities. “We published a report in 2012 where we estimated 22 veterans a day die by suicide and that caught fire, but it is purely an estimate. The truth is we have no idea what the real count is, because until we began working on the Suicide Data Repository there was no national register of veteran mortality and there has been no way for us to know.”

Dr. Bossarte anticipates that the VA will have the data necessary to calculate more accurate statistics by the end of the year. “More than 70% of veteran suicide are in people over age 50; but the rates are going up most among the youngest.”

A notable drop in veteran suicide rates for those who used services occurred between 2001 and 2003, and that decrease remains unexplained; it preceded later changes in mental health services and enhanced suicide prevention programs. Dr. Bossarte also pointed out that just under half of veterans who die from suicide have no mental health diagnosis, despite yearly screening to identify people who may be suffering from posttraumatic stress disorder, alcohol-related disorders, and depression.

“The attention on veteran suicide started around 2007,” Dr. Bossarte explained. “Mark Kaplan published a study using publicly available mortality data; those who reported they were veterans were twice as likely to die of suicide as those who were not.”

While this sparked interest in veteran suicide, it’s important to note that a replication of that study in 2012 by Matthew Miller did not have the same findings.

“Then, in 2008, for the first time in recent history,” Dr. Bossarte continued, “the suicide rate among active duty military personnel exceeded that of the general population. Traditionally, rates of suicide in this population have been 40%-50% lower than in the general population. The increased rate was seen primarily in the Army and Marines. Serious mental illness may make people ineligible for military service, as can violent and disruptive behavior – things that are associated with suicide – so you tended to get a healthier population in the military.”

Dr. Bossarte noted that there was conjecture that increased suicide rates among active military might be related to more waivers that allowed people to enlist who would not ordinarily be eligible, and/or to higher rates of deployment. He went on to talk about Army STARRS (Army Study to Assess Risk and Resilience in Servicemembers).

“STARRS devoted $50 million over 5 years to the largest suicide study and did not find an effect of waivers. They did report a higher suicide rate among those who were deployed, however. But then Tim Bullman in my office looked at suicide rates 7 years after separation from service, and he reported a higher suicide rate among those who were never deployed.” The VA studies, I quickly realized, were also confusing and contradictory.

The VA has greatly expanded its mental health and suicide prevention services. For veterans overall, suicide rates have stabilized, but they have not decreased. For those veterans with psychiatric disorders, however, the suicide rates have gone down.

“When you ask ‘does treatment matter?’ it’s so hard to disentangle psychotherapy from pharmacotherapy. Over the past decade, we’ve seen a significant decrease in the suicide rate among those veterans with mental health disorders. We’ve looked at suicide rates every way you can think of. One thing we do know is that the better the relationship with the clinician, the lower the suicide risk.”

We talked about the role of hospitalization in preventing suicide, and Dr. Bossarte noted that the highest risk for suicide is immediately following hospital discharge.

“We are looking at people hospitalized after their first-ever suicide attempts and rates of mortality, including suicidal behavior, for 1 year after discharge. In very preliminary findings, we didn’t see any difference in the outcome for either all-cause mortality or repeat suicide attempts in those who were hospitalized, compared to those who were not. We don’t yet know about completed suicide.”

I left my discussion with Dr. Bossarte with more questions than answers. We have reason to believe that treatment helps, but we still don’t know which treatments help which people, and we do know that treatment doesn’t prevent suicide in every patient. In a culture where “treatment” has come to be equated with “prescribing” and is often based on a checklist of symptoms done by a primary care clinician, one might wonder if combining psychotherapy and medication – an increasingly rare offering – might have a better outcome. Simply put, for a problem that prematurely takes more than 40,000 lives a year, we know much too little.

Dr. Miller is a coauthor of “Shrink Rap: Three Psychiatrists Explain Their Work” (Baltimore: Johns Hopkins University Press, 2011).

Over the past 2 decades, more and more people have been treated with antidepressant medications. In the same period of time, suicide rates have gone up – not down. To those of us who treat patients, this fact is both surprising and perplexing. It seems that suicidal thoughts are a common feature of major depression, and when the depressive symptoms abate with treatment, the suicidal thoughts dissipate. Intuitively, it seems that treating depression on a larger scale should prevent suicides, but we still don’t know that conclusively.

According to the National Health and Nutrition Examination Survey (NHANES), 11% percent of Americans over the age of 12 are taking an antidepressant medication. In women aged 40-59, this number is 23%. Of those taking antidepressants, only one-third have seen a mental health professional in the past 12 months. What also is striking is that for people surveyed with symptoms of severe depression, only one-third were on medication.

In 2013, just over 40,000 Americans died of suicide. From 2000 to 2013, the suicide rate per 100,000 Americans has steadily increased from 10.4 to 12.6 per 100,000 people. While we know that people with psychiatric illnesses have higher rates of suicide compared with the general population, what we don’t know is whether the people dying are the same people who are getting treatment.

Thinking about this gets very difficult. It has been estimated that 90% of those who die of suicide have suffered from a mental illness. This figure includes those who were treated, untreated, and previously treated, but the studies have methodologic inconsistencies and that 90% estimate may not be accurate. Certainly, however, people die of suicide for reasons that have nothing to do with psychiatric illness, and we do know that impulsive responses to distressing circumstances are a factor, especially when a lethal method is easily available.

Several studies have shown that antidepressant use, particularly in older adults, may be associated with a decrease in suicidality. The studies often combine suicide attempts and completions. The issues with younger patients are more complicated, and in recent weeks, the reanalysis of the 2001 Paxil 329 study has again raised issues about the safety of certain antidepressants in children and adolescents. The data for all these studies are both confusing and contradictory, and are not easy to examine or interpret.

We also don’t know what role psychotherapy plays. A study done at Johns Hopkins Bloomberg School of Public Health looked at the follow-up for 65,000 people in Denmark who had attempted suicide and found that the rates for completed suicide dropped if the patient received a short course of psychosocial therapy at a suicide prevention center. But again, this study looked at a select group of people who had already attempted suicide.

I began to think it might help to ask these questions in a closed system where patients could be tallied with regard to requests for treatment, what type of treatment was provided, and even access to autopsy results. The U.S. Department of Veterans Affairs seemed to be a source where such answers might be found. It has been reported that 22 veterans a day die by suicide, and many veterans get their care in VA facilities, with VA pharmacy benefits, and treatment effects can, in theory, be studied.

Hoping to get a better sense of the relationship between treatment and suicide, I met with Robert Bossarte, Ph.D., director of the Epidemiology Program in the VA’s Office of Public Health. His career has been focused on suicide prevention.

The first thing Dr. Bossarte did was dissuade me of the idea that the VA is a closed system. Not all veterans receive lifelong benefits from the VA, and the formulas for determining who is entitled to what benefits, and for how long, is rather complicated. Dr. Bossarte also noted that some patients go outside of the system for their care.

“What we do know is that among those who have used VA services in the previous year, about 2,000 veterans a year die from suicide. It’s been hovering around that for the past decade,” Dr. Bossarte explained, again emphasizing that many veterans do not receive care at VA facilities. “We published a report in 2012 where we estimated 22 veterans a day die by suicide and that caught fire, but it is purely an estimate. The truth is we have no idea what the real count is, because until we began working on the Suicide Data Repository there was no national register of veteran mortality and there has been no way for us to know.”

Dr. Bossarte anticipates that the VA will have the data necessary to calculate more accurate statistics by the end of the year. “More than 70% of veteran suicide are in people over age 50; but the rates are going up most among the youngest.”

A notable drop in veteran suicide rates for those who used services occurred between 2001 and 2003, and that decrease remains unexplained; it preceded later changes in mental health services and enhanced suicide prevention programs. Dr. Bossarte also pointed out that just under half of veterans who die from suicide have no mental health diagnosis, despite yearly screening to identify people who may be suffering from posttraumatic stress disorder, alcohol-related disorders, and depression.

“The attention on veteran suicide started around 2007,” Dr. Bossarte explained. “Mark Kaplan published a study using publicly available mortality data; those who reported they were veterans were twice as likely to die of suicide as those who were not.”

While this sparked interest in veteran suicide, it’s important to note that a replication of that study in 2012 by Matthew Miller did not have the same findings.

“Then, in 2008, for the first time in recent history,” Dr. Bossarte continued, “the suicide rate among active duty military personnel exceeded that of the general population. Traditionally, rates of suicide in this population have been 40%-50% lower than in the general population. The increased rate was seen primarily in the Army and Marines. Serious mental illness may make people ineligible for military service, as can violent and disruptive behavior – things that are associated with suicide – so you tended to get a healthier population in the military.”

Dr. Bossarte noted that there was conjecture that increased suicide rates among active military might be related to more waivers that allowed people to enlist who would not ordinarily be eligible, and/or to higher rates of deployment. He went on to talk about Army STARRS (Army Study to Assess Risk and Resilience in Servicemembers).

“STARRS devoted $50 million over 5 years to the largest suicide study and did not find an effect of waivers. They did report a higher suicide rate among those who were deployed, however. But then Tim Bullman in my office looked at suicide rates 7 years after separation from service, and he reported a higher suicide rate among those who were never deployed.” The VA studies, I quickly realized, were also confusing and contradictory.

The VA has greatly expanded its mental health and suicide prevention services. For veterans overall, suicide rates have stabilized, but they have not decreased. For those veterans with psychiatric disorders, however, the suicide rates have gone down.

“When you ask ‘does treatment matter?’ it’s so hard to disentangle psychotherapy from pharmacotherapy. Over the past decade, we’ve seen a significant decrease in the suicide rate among those veterans with mental health disorders. We’ve looked at suicide rates every way you can think of. One thing we do know is that the better the relationship with the clinician, the lower the suicide risk.”

We talked about the role of hospitalization in preventing suicide, and Dr. Bossarte noted that the highest risk for suicide is immediately following hospital discharge.

“We are looking at people hospitalized after their first-ever suicide attempts and rates of mortality, including suicidal behavior, for 1 year after discharge. In very preliminary findings, we didn’t see any difference in the outcome for either all-cause mortality or repeat suicide attempts in those who were hospitalized, compared to those who were not. We don’t yet know about completed suicide.”

I left my discussion with Dr. Bossarte with more questions than answers. We have reason to believe that treatment helps, but we still don’t know which treatments help which people, and we do know that treatment doesn’t prevent suicide in every patient. In a culture where “treatment” has come to be equated with “prescribing” and is often based on a checklist of symptoms done by a primary care clinician, one might wonder if combining psychotherapy and medication – an increasingly rare offering – might have a better outcome. Simply put, for a problem that prematurely takes more than 40,000 lives a year, we know much too little.

Dr. Miller is a coauthor of “Shrink Rap: Three Psychiatrists Explain Their Work” (Baltimore: Johns Hopkins University Press, 2011).