Switching Antidepressants May Boost Suicide Risk in Elderly

Deborah Brauser

March 28, 2018

HONOLULU, Hawaii — Prescribing patterns for antidepressant treatment can play a part in increasing risk for suicidal behavior in elderly patients, new research suggests.

The population-based cohort study, which included more than 185,000 Swedish residents aged 75 years or older who initiated antidepressant treatment late in life, showed that switching to another antidepressant more than doubled the risk for suicide and almost doubled the risk for suicide attempts.

Risk for both suicide attempts and completions was also significantly increased for those using anxiolytics or hypnotics at the same time as an antidepressant.

"We identified three prescription patterns as being associated with both suicide and suicide attempt, which are things that prescribers should at least keep in mind," Margda Waern, MD, PhD, professor and senior psychiatrist at the Institute of Neuroscience and Physiology at the University of Gothenburg, Sweden, told Medscape Medical News.

"If you're considering switching medications in this population, do it very slowly," added Waern.

Interestingly, concomitant use of antidementia drugs was associated with a decreased risk for suicide attempts.

The findings, which were published last month in the European Journal of Clinical Pharmacology, were presented and discussed here at American Association for Geriatric Psychiatry (AAGP) 2018.

Few Studies on Use Patterns

One fifth of the Swedish population who are older than 75 years take antidepressants, Waern told meeting attendees.

"Yet few studies investigate how different use patterns may impact on the risk of suicidal behavior in older adults," she said. "The identification of depression is very important as a strategy for suicide prevention, but antidepressants won't always do the trick."

The investigators examined data on 185,225 patients at least 75 years of age (mean age, 83.4 years; 63.5% women) who filled a prescription for an antidepressant between January 2007 and December 2013. Follow-up extended through 2014.

Selective serotonin-reuptake inhibitors (SSRIs) were used by 63.1% of the participants. The most commonly used SSRIs were citalopram (multiple brands) (48.4%), sertraline (Zoloft, Pfizer) (9.9%), and escitalopram (Lexapro, Allergan) (3.3%). SNRIs/norepinephrine reuptake inhibitors and tricyclic antidepressants were used by 25.3% and 11.6%, respectively.

Prescription patterns showed that 14.8% of the patients were using two or more antidepressants; 50.8% concomitantly used other psychiatric medications. Of this latter group, 32% used hypnotics, 19.9% used anxiolytics, 7.6% used antipsychotics, and 5.8% used antidementia drugs.

During the follow-up period, there were 295 completed suicides (incidence rate, 50 per 100,000 person-years for total study population) and 654 suicide attempts (incidence rate, 117 per 100,000 person-years).

The incidence rates of suicides and suicide attempts for women were 25 and 94 per 100,000 person-years, respectively. For men, the rates were 106 and 167 per 100,000 person-years.

The adjusted subhazard ratios (SHRs) for the study participants who switched to another antidepressant treatment vs those who did not switch were 2.42 for suicides (95% confidence interval [CI], 1.65 - 3.55; P < .001) and 1.76 for suicide attempts (95% CI, 1.32 - 2.34; P < .001).

In addition, those who concomitantly filled prescriptions for hypnotics were at significantly increased risk for both suicide and suicide attempts, as were those who filled prescriptions for anxiolytics, as shown in the table below.

Table. Other Risk Factors for Suicide, Suicide Attempts

Drugs Used Concomitantly Adjusted SHR for Suicide (95% CI) P Value Adjusted SHR for Suicide Attempt (95% CI) P Value
Hypnotics 2.20 (1.69 - 2.85 < .001 2.86 (2.38 - 3.43) < .001
Anxiolytics 1.54 (1.20 - 1.96) < .001 2.04 (1.73 - 2.40) < .001


The risk for suicide was also significantly increased for those who also used antipsychotics (SHR, 1.73; P = .004). However, the risk for suicide attempts was not significantly increased (SHR, 1.16), making the overall association with the use of antipsychotics "inconsistent," write the investigators.

In those who were taking antidementia drugs along with antidepressants, the risk for suicide attempt was significantly decreased (SHR, 0.40; 95% CI, 0.27 - 0.59; P < .001).

"It's difficult to know what's going on with this finding," admitted Waern. "Some of these people might have been in a nursing home setting. And that might have been the thing that was decreasing their risk, having people around them."

Although more research is needed to draw definite conclusions, the overall study findings "may help to inform the prescribers who initiate antidepressant treatment in their older patients," write the investigators.

"We can't say from this type of registry study what the mechanism is for what's going on. But we can certainly say: if you're considering switching, know that this could be a marker of increased risk for both suicide and suicide attempt. So be very careful with the older patient," said Waern.

Suicide in the Elderly an "Epidemic"

"We're in the middle of an epidemic right now" regarding suicide in the elderly, session moderator Gary J. Kennedy, MD, Montefiore Medical Center at Albert Einstein College of Medicine, New York City, told Medscape Medical News.

He added that the information from Waern's observation was interesting and clinically useful.

"Maybe what she's seeing is severity of mental illness. These people are switching from one medication to another because they aren't satisfied or it isn't working," said Kennedy.

"From a population base, if we track people a little more carefully with their medication use, perhaps we could reduce their suicide risk," he said.

However, he noted that the Swedish registries, which track universal care, differ from electronic records used in the United States, where it's harder to track information from one hospital to another. "It's a real problem here."

Earlier in the session, Kennedy presented new data in which talk of being "better off dead" or of self-harm was distinguished from explicit suicidal ideation.

Risk Factors

The New York City Neighborhood and Mental Health Study (NYCNAMES II) was created to determine risk for depression, as well as neighborhood characteristics and physical activity patterns, in a community-based cohort of adults aged 65 to 75 years.

Participants who answered positively to the ninth item on the Patient Health Questionnaire (PHQ-9, n = 222) were called. Of those who completed a psychiatric assessment, 82 were not experiencing sucidal ideation, and 60 were.

Risk assessment outcomes showed that illness and economic stress were factors (P = .01 and < .05, respectively) that could distinguish patients experiencing suicidal ideation from those who were not. Interestingly, severity of depression was not a distinguishing factor.

"Simply asking them about their health may be a better way than asking about mood for determining risk," said Kennedy. "To my surprise, pain, disability, family, and bereavement were not associated with suicidal ideation," he added.

The best intervention for mitigating risk was counseling (P = .0001). Kennedy noted that one response he received as a reason to live was, "I'd never abandon my dog."

"So in an apartment-type place like Manhattan, it's important to ask about the pets. I tell my residents and medical students: always ask the pet's name," he said.

"The [overall] findings from NYCNAMES II suggest that the ninth item of the PHQ-9 is an inexact proxy query for suicidal ideation. And addressing the older adult's thoughts on suicide may require more than simply focusing on depression," he summarized.

The first study was funded by grants from the Swedish Research Council; the Swedish Research Council for Health, Working Life, and Welfare; and the Sönderstrom-König Foundation. Dr Waern has disclosed no relevant financial relationships. Dr Kennedy has received honoraria from the Guilford Press.

American Association for Geriatric Psychiatry (AAGP) 2018 Annual Meeting. Session 323, presented March 17, 2018.

Eur J Clin Pharmacol. 2018;74:201-208. Full article

Follow Deborah Brauser on Twitter: @MedscapeDeb. For more Medscape Psychiatry news, join us on Facebook and Twitter.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.