Antidepressants in Older Patients May Have Adverse Outcomes

Medscape |

Clinical Context

According to the January 25, 2006, issue of the Cochrane Database Systematic Review by Mottram and colleagues of elderly people with depression, tricyclic antidepressants (TCAs) and selective serotonin reuptake inhibitors (SSRIs) have similar efficacy, but TCAs appear to have a higher discontinuation rate because of adverse effects. In 2009, the National Institute for Health and Clinical Excellence National Clinical Practice Guideline 90 recommended SSRI use for treatment of depression, but adverse effects and patient preference should also be considered.

This cohort study by Coupland and colleagues assesses whether antidepressant class or individual antidepressants are associated with certain adverse outcomes in elderly persons with depression.

Study Synopsis and Perspective

Use of antidepressant medication in those over age 65 is risky, new research suggests.

A population-based cohort study in the United Kingdom published online August 2 in the British Medical Journal showed significant associations between the use of antidepressants and adverse outcomes, including falls, stroke, seizures, and all-cause mortality in elderly people with depression.

Patients prescribed SSRIs, which were the most commonly prescribed antidepressants, fared worse than those receiving the older tricyclic antidepressants, according to the researchers.

“Our findings are new and unexpected, and need to be confirmed in other studies,” lead author Carol Coupland, PhD, University of Nottingham, United Kingdom, told Medscape Medical News.

“The benefits of the different classes of antidepressant drugs also need to be considered alongside the adverse effects,” Dr. Coupland added. “If our findings are confirmed, then low-dose TCAs should be considered when assessing antidepressant treatment for older people with depression.”

Under-Represented Group

The study was commissioned and funded by the United Kingdom’s National Institute of Research Health Technology Assessment program in 2006 to investigate the potential risks of antidepressant use in the elderly.

“Little is known about the adverse effects associated with antidepressants in the elderly, who may be at increased risk because of their higher levels of comorbidity, age-related physiological changes, and polypharmacy,” Dr. Coupland said.

“Older people are also under-represented in clinical trials of antidepressants, and most of these trials are short-term, which makes it difficult to derive reliable estimates of the incidence of adverse events in this group,” she said.

Dr. Coupland and her team identified 60,746 patients aged 65 to 100 years with a newly diagnosed episode of depression between 1996 and 2007. The mean age of the patients was 75 years; 33.3% of patients were men and 66.7% were women. Many patients had other conditions, such as heart disease and diabetes, and were taking several medications.

They were followed for a mean of 5 years, and the total number of person-years of follow-up was 305,188.


During follow-up, 89% of the cohort (n = 54,038) received at least 1 prescription for an antidepressant drug. SSRIs accounted for 54.7%, TCAs made up 31.6%, and the group of other antidepressants (venlafaxine hydrochloride, and mirtazapine) made up 13.5% of prescriptions. Only 0.2% of prescriptions were for monoamine oxidase inhibitors, and patients who were prescribed these drugs were excluded from the analysis.

After adjusting for confounding factors, including age, sex, severity of depression, other illnesses, and use of other medications, the study showed that SSRIs and drugs in the group of other antidepressants were associated with an increased risk for several adverse outcomes compared with TCAs.

For the SSRIs, the adjusted hazard ratios (HR) were as follows:

  • All-cause mortality: 1.32 (95% confidence interval [CI], 1.26 – 1.39)
  • Stroke or transient ischemic attack: 1.15 (95% CI, 1.05 – 1.26)
  • Falls: 1.27 (95% CI, 1.20 – 1.35)
  • Fracture: 1.26 (95% CI, 1.15 – 1.37)
  • Epilepsy or seizures: 1.80 (95% CI, 1.32 – 2.43)
  • Hyponatremia: 1.44 (95% CI, 1.19 – 1.75)

For the group of other antidepressants the adjusted HRs were:

  • All-cause mortality: 1.43 (95% CI, 1.33 – 1.54)
  • Attempted suicide or self-harm: 3.04 (95% CI, 2.21 – 4.17)
  • Stroke or transient ischemic attack: 1.35 (95% CI, 1.18 – 1.54)
  • Fracture: 1.31 (95% CI, 1.15 – 1.50)
  • Epilepsy or seizures: 2.20 (95% CI, 1.46 – 3.30)

The analysis also showed SSRIs were associated with the highest adjusted HR for falls (HR, 1.66; 95% CI, 1.58 – 1.73) and hyponatremia (HR, 1.52; 95% CI, 1.33 – 1.75) vs no antidepressant use.

Similarly, in comparison with no antidepressant use, other classes of antidepressants were associated with the highest adjusted HRs for all-cause mortality (1.66; 95% CI, 1.56 – 1.77), attempted suicide or self-harm (5.16; 95% CI, 3.90 – 6.83), stroke or transient ischemic attack (HR, 1.37; 95% CI, 1.22 – 1.55), fracture (HR, 1.64; 95% CI, 1.46 – 1.84), and epilepsy or seizures (HR, 2.24; 95% CI, 1.60 – 3.15).

Among individual drugs, trazodone, mirtazapine and venlafaxine were associated with the highest risks for several outcomes.

TCAs Less Risky?

The absolute risks over 1 year for all-cause mortality were 7.04% for patients, 8.12% for TCAs, 10.61% for SSRIs, and 11.43% for other antidepressants vs no antidepressant use.

The most dangerous times for adverse events were within the first 28 days of starting an antidepressant and within the first 28 days after stopping.

“This is new research and needs confirmation in further studies. It has shown associations but not necessarily a direct causal link, so there is no need for patients to change or stop their treatments at this stage if they are finding them beneficial,” Dr. Coupland cautioned.

“Doctors should also advise patients that adverse effects are most common during the first few weeks of treatment, and they should be monitored during this period,” she added.

Asked whether TCAs are safer to use in older individuals, Dr. Coupland replied: “Our study suggests this might be the case, but our findings could also be due to residual confounding due to differences between patients prescribed different antidepressant drugs.”

Her hope is that this study will prompt further research leading to clearer guidance on how best to treat depression effectively but safely in older people.

Findings Inform Prescribing

In an accompanying editorial, Ian B. Hickie, MD, from the University of Sydney in Australia, writes that the study has clear implications for more informed prescribing and enhanced clinical monitoring.

“Given the potential harms, the decision to prescribe for an older person with depression should not be taken lightly,” he writes.

He adds that older people require careful monitoring for adverse effects and recommends that they be seen at least weekly during their first month of treatment and again when the drugs are stopped.

Medscape Medical News asked Josepha A. Cheong, MD, professor of psychiatry at the University of Florida College of Medicine in Gainesville, to comment on the findings.

Dr. Cheong, who specializes in geriatric psychiatry, said the paper was important because it serves as a general reminder that geriatric patients are more likely to have medical comorbid conditions, medical complications, and increased sensitivity to side effects at lower doses than younger matched cohorts.

She also questioned the suggestion that TCAs may be safer in older people.

“Just because traditional TCAs in low doses were found to have fewer associated mortality and morbidity outcomes than SSRIs does not necessarily imply that they are safer or better tolerated, as their inherent side effect profiles tend to decrease efficacy due to patients’ early discontinuation or intolerance of side effects at therapeutic doses,” she said.

Dr. Cheong agreed it is important to monitor geriatric patients very closely.

“Use caution whenever starting any medication in a geriatric patient,” she said. “Carefully consider concurrent medications and potential for drug-drug interactions, prescribe a medication only if absolutely indicated and with clear therapeutic goals that can be documented with assessment, and use the minimum therapeutically effective dose to maximize therapeutic response while minimizing potential side effects.”

This study was funded by NIHR Health Technology Assessment Programme. Dr. Coupland has disclosed no relevant financial relationships. Dr. Hickie reports that his research is supported principally by a National Health and Medical Research Council Australian Medical Research Fellowship and that he has also has participated in depression awareness projects for health professionals supported by government, community agency, and industry sponsors, including Wyeth, Eli Lilly, Servier, Pfizer, and AstraZeneca. Dr. Cheong has disclosed no relevant financial relationships.

BMJ. Published online August 2, 2011.

Related Link
A 2006 Cochrane review on use of antidepressants for depressed elderly is available online.

Study Highlights


  • 60,746 patients aged between 65 and 100 years old with a new episode of depression during a 12-year period were enrolled.
  • 570 general practices in the United Kingdom contributed to a primary care database.
  • Exclusion criteria were age 100 years or older, diagnosis less than 12 months after registration with a study practice or practice’s installation of a computer system, temporary residence, depression diagnosis or antidepressant prescription in the 12 months before the diagnosis was recorded, and psychosis.
  • Mean age of the patients was 75.0 years.
  • 20,230 (33.3%) were men, and 40,516 (66.7%) were women.
  • Mean follow-up period was 5.0 years per person.
  • 54,038 (89.0%) received at least 1 prescription for antidepressant during follow-up.
  • 1,398,359 antidepressant prescriptions were provided:
    • 764,659 SSRIs (54.7%)
    • 442,192 TCAs (31.6%)
    • 2203 monoamine oxidase inhibitors (0.2%)
    • 189,305 other antidepressants (13.5%)
  • The low number of prescriptions for monoamine oxidase inhibitors precluded use of data.
  • Data on adverse outcomes were obtained from computer records and death certificates for a 13-year study period that included 1 year after the last year of patient enrollment.
  • The primary outcome measures were HRs for all-cause mortality, attempted suicide or self-harm, myocardial infarction, stroke/transient ischemic attack, falls, fractures, upper gastrointestinal tract bleeding, epilepsy/seizures, road traffic crashes, adverse drug reactions, and hyponatremia.
  • Analysis adjusted for age at study entry, sex, year of depression diagnosis, previous depression diagnosis before age 65 years, the severity of index diagnosis of depression, deprivation, smoking status, comorbidities, use of other drugs at baseline, and previous falls before baseline.
  • SSRIs vs no treatment were linked with the highest adjusted HR for falls (HR, 1.66) and hyponatremia (HR, 1.52).
  • Other antidepressants vs no treatment were linked with the highest adjusted HR for all-cause mortality (HR, 1.66), attempted suicide/self-harm (HR, 5.16), stroke/transient ischemic attack (HR, 1.37), fracture (HR, 1.64), and epilepsy/seizures (HR, 2.24).
  • TCAs vs no treatment did not have the highest HR for any outcome.
  • SSRIs vs TCAs had a higher adjusted HR for all-cause mortality (HR, 1.32), stroke/transient ischemic attack (HR, 1.15), falls (HR, 1.27), fracture (HR, 1.26), epilepsy/seizures (HR, 1.80), and hyponatremia (HR, 1.44).
  • Other antidepressants vs TCAs had a higher adjusted HR for all-cause mortality (HR, 1.43), attempted suicide/self harm (HR, 3.04), stroke/transient ischemic attack (HR, 1.35), fracture (HR, 1.31), and epilepsy/seizures (HR, 2.20).
  • TCAs vs SSRIs or other antidepressants did not have higher rates of any adverse outcomes.
  • The most commonly prescribed individual medications were citalopram, fluoxetine, paroxetine, and sertraline (SSRIs); amitriptyline, dosulepin, lofepramine, and trazodone (TCAs); and venlafaxine and mirtazapine (other).
  • Individual medications had the highest HR for certain adverse outcomes:
    • Trazodone (TCA): all-cause mortality
    • Mirtazapine (other): attempted suicide/self-harm
    • Venlafaxine (other): stroke/transient ischemic attack, fracture, and epilepsy/seizures
    • Citalopram (similar to other SSRIs): falls
    • Citalopram, escitalopram, and fluoxetine: hyponatremia
  • Dose trends were noted for TCAs and SSRIs for all-cause mortality, falls, and epilepsy/seizures and for TCA with fracture.
  • Most outcomes had the highest risks in the first 28 days after initiation of medication and after discontinuation of medication.
  • Absolute risks for all-cause mortality at 1 year were 7.04% for no antidepressant use, 8.12% for TCA use, 10.61% for SSRI use, and 11.43% for other antidepressants.
  • Study limitations included observational design and residual confounding.

Clinical Implications


  • In elderly patients with depression, SSRIs vs TCAs are linked with higher rates of all-cause mortality, stroke/transient ischemic attack, falls, fracture, epilepsy/seizures, and hyponatremia. Other antidepressants vs TCAs are linked with higher rates of all-cause mortality, attempted suicide/self-harm, stroke/transient ischemic attack, fracture, and epilepsy/seizures.
  • In elderly patients with depression, individual antidepressant drugs linked with the highest risk for certain adverse outcomes include the TCA trazodone for all-cause mortality; mirtazapine for attempted suicide/self-harm; venlafaxine for stroke/transient ischemic attack, fracture, and epilepsy/seizures; and citalopram, escitalopram, and fluoxetine for hyponatremia.

Leave a Reply