Patients who are prescribed citalopram or nortriptyline are at increased risk of sudden cardiac death, a Danish study has found.

A total of 19,110 patients who had suffered a cardiac arrest outside hospital (OHCA) were identified from the Danish Cardiac Arrest Register. Of these, 2,913 (15.2% of the sample) were receiving antidepressant treatment at the time of their arrest. An age- and sex-matched control group was identified from the entire population.

A case time control design was used. For each subject the case window was 30 days prior to the cardiac arrest and the 30 day control periods were 60-90 and 90-120 days prior to the arrest.

The case-crossover odds ratio is the ratio of the odds of antidepressant exposure during the case and control periods. The control group is included to remove bias due to changes in antidepressant prescribing over time. The case time control ratio is the case-crossover odds ratio divided by the time trend odds ratio.¹

This complex design reduces the risk of confounding as the cases act as their own controls. It estimates whether the risk of cardiac arrest is increased in patients who have recently commenced antidepressant treatment.

A significant association was found between OHCA and the initiation of both tricyclic and SSRI antidepressants: odds ratios (OR) 1.69 (95% CI: 1.14-2.50) and 1.21 (95% CI: 1.00-1.49) respectively. There was no significant association with initiation of serotonin-norepinephrine reuptake inibitors (SNRIs)/noradrenergic and specific serotonergic antidepressants i.e. venlafaxine, mianserin and mirtazapine (NaSSAs).

When antidepressants were considered individually, a significant association was found only for citalopram OR 1.29 (95% CI: 1.02-1.63) and nortriptyline OR 5.14 (95% CI: 2.17-12.2).

Citalopram was by far the most frequently prescribed antidepressant (50.8%). Relatively few patients were prescribed other SSRIs, so the confidence intervals are wide and we cannot therefore be certain that these drugs are safe. However, the results are consistent with other evidence to suggest that citalopram is proarrhythmic. A UK observational study found that citalopram was more likely than other SSRIs to prove fatal in overdose,2 and the MHRA has issued a warning that citalopram and escitalopram prolong the QT interval and have been linked to Torsade de Pointes.³ Conversely, the SADHART study found that sertraline is a safe treatment following myocardial infarction.⁴

Tricyclics are also proarrhythmic and increase the QT interval. Therefore, it is surprising that only nortriptyline was associated with OHCA. The authors suggest that the lack of association between amitriptyline and OHCA may have been because amitriptyline was prescribed in low doses for the treatment of neuropathic pain.

In 2010, citalopram was the most frequently prescribed antidepressant in England and Wales (11.8 and 1.1 million items respectively).⁵ Although it remains a good choice of drug for many patients, in my opinion it should no longer be considered the antidepressant of first preference, particularly for older patients and those with cardiac disease.

Dr Phillip Bland

• Weeke P, Jensen A, Folke F et al. Antidepressant use and risk of out-of-hospital cardiac arrest: a nationwide case-time-control study. Clin Pharmacol Ther 2012;92:72-9