Initiation of Treatment
On average, all antidepressants are equally effective. Although an individual patient may preferentially respond to a certain anti-depressant, it is difficult to predict this in advance. Without a personal or family history of such a response, side effects are the most influential factors in choosing an agent. Side effects may be particularly relevant in the following groups of patients.
There is a high rate of depression in patients who have had a myocardial infarction and the presence of depression adversely affects the prognosis of cardiac disease.
The major cardiovascular side effect of TCAs is orthos-tatic hypotension. This can be clinically significant in both the hypertensive patient and the elderly patient. Some of the tricyclic medications may have a lower risk of orthostatic hypotension, notably nortriptyline and doxepin. The evidence for nortriptyline causing little or no hypotension is convincing; for doxepin it is weaker (Roose and Dalack, 1992).
TCAs do not show a negative inotropic effect and do not seem to worsen congestive heart failure. Patients with congestive heart failure may, however, be at a higher risk for orthostatic hy-potension. Again, nortriptyline is the safest TCA in this case.
TCAs slow conduction at the bundle of His. Their effect is analogous to type 1A antiarrhythmic agents such as quinidine and procainamide. In therapeutic doses they can slow cardiac conduction and in overdose they can cause atrioventricular block-ade. These effects are of most concern in patients with preexist-ing cardiac conduction defects. There is no evidence that any one tricyclic medication is safer than another.
Glassman and coworkers (1993) recommend caution when using TCAs in all patients with ischemic heart disease, particularly patients with ventricular arrhythmias that follow a myocardial infarction.
The serotonin reuptake inhibitors differ from the TCAs in that they do not prolong the PR or heart wave (QRS) interval. Thus, they probably lack any of the proarrhythmic and antiarrhythmic activities associated with TCAs. They do not cause orthostatic hypotension. Most studies that exist suggest that SSRIs should be safe in patients with heart disease.
Trazodone has less proarrhythmic effect than the TCAs. However, there have been reports of trazodone-related ventricular ectopy and complete heart block (Martyn et al., 1993). In patients with preexisting heart disease, bupropion does not appear to cause the cardiac side effects attributed to TCAs (Roose et al., 1991).
Two pharmacokinetic changes are of great importance in ag-ing patients: decreased efficiency of the hepatic microoxidase system and a decreased muscle–fat ratio. Decreased efficiency of hepatic microoxidases results in the slower metabolism of antidepressants and other drugs. Normal increases in body fat and a loss of muscle mass result in an alteration of the volume of distribution for a substance. Thus, lipophilic drugs, including all antidepressants, are more widely distributed in the elderly body.
Both the resulting slower metabolism and the increased volume of distribution increase the half-lives of the various anti-depressants. The elderly, therefore, are likely to have a greater incidence of side effects.
The half-lives and steady state concentrations of the se-rotonin reuptake inhibitors are only minimally affected by age. Paroxetine may be an exception to this, and it may have an in-creased half-life in the elderly (Leonard, 1993).
The assumption that elderly patients require lower blood levels of antidepressant medication is not correct. Although pharma-cokinetic concerns may require lowering of the medication dose, plasma levels are comparable to those in young adults.
It appears that antidepressants are as effective in the eld-erly as in the young, although the elderly may have more diffi-culty tolerating these medications.
Few studies are available on the pharmacological treatment of depression in medically ill patients, and even less on medically ill elderly patients. In initiating treatment, the physician should understand the effects of various illnesses on pharmacokinetics and the potential side effects that may result.
Next to stimulants, antidepressants are the most common psy-chotropic medications prescribed in children (Zito et al., 2000) and adolescents (Jensen et al., 1999) and the trend to use these medications in that group is on an increase. It should be noted that the FDA has placed a “black box” warning related to suicidal ideation in adolescents treated with SSRIs.
In addition to depression, antidepressants are used in a number of nonaffective disorders in children and adolescents, including enuresis, attention-deficit/hyperactivity disorder, OCD and bulimia nervosa. They are used to mitigate behaviors that are not part of a disease, for example, in suicidal adolescents who do not have a concomitant depressive disorder (American Academy of Child and Adolescent Psychiatry, 2001); however, there is little empirical evidence that suggests that they can lower suicide risk (Zametkin et al., 2001). Similarly, they are used to dampen other impulsive behaviors.
Children should generally be treated with levels of medication comparable to those used in adults, with the doses adjusted for body weight. As children have large livers (relative to body weight), they tend to be efficient metabolizers of substances, and may even need higher doses of medications (relative to body weight) than do adults.
Well after the introduction of SSRIs, TCAs remained the most popular antidepressants used in children. However, there have been concerns regarding the safety of these medications. Sev-eral case studies have documented adverse cardiac affects in children treated with TCAs, especially with desipramine. The Work Group on Research of the American Academy of Child and Adolescent Psychiatry (1990) conducted an exten-sive review of the available literature and concluded that there is minimal or no increased risk of “sudden death” in children with desipramine. It is probably wise, however, to perform an adequate cardiac work-up before starting an antidepressant in a child.
Given these issues concerning cardiotoxicity, the serot-onin reuptake inhibitors have become more popular for the treat-ment of children and adolescents. Most of the newer agents have evidence for safety and efficacy in children, including fluoxetine (Emslie et al., 1997) and paroxetine (Keller et al., 2001). Others of the SSRIs have data for their usefulness in anxiety disorders, such as fluvoxamine (in OCD) (Riddle et al., 2001) and sertra-line (which has open label data for social anxiety disorder and OCD) (Compton et al., 2001). Bupropion has been found safe and effective in children as well, although it may exacerbate tics in Tourette’s disorder. Other antidepressants, including most of the first-generation antidepressants, lack any positive efficacy stud-ies in children or adolescents.
Of the commonly used antidepressants, none show compel-ling evidence of teratogenicity. Nonetheless, as is true for most medications, it is usually best to discontinue antidepressants if a patient becomes pregnant. For all the antidepressants, one must consider a risk–benefit decision, measuring the risk of possible pregnancy-related side effects against the morbidity of a pro-longed untreated depressive episode during pregnancy (Wisner et al., 2000). The FDA, in late 2005, classified paroxetine as a class “D” medication, indicating it should be used only with great caution in pregnancy.
All antidepressants are secreted in breast milk. The levels of these agents are difficult to predict, and their effect on the de-veloping infant is not known (Cohen et al., 1991). There are no systematic investigations of this issue. Given the level of uncer-tainty, nursing should be deferred if antidepressant treatment is required.
Meta-analysis of both world (Mann and Kapur, 1991) and the USA (Kapur et al., 1992) and a more recent study using prospec-tive data (Leon et al., 1999) showed no unique relationship be-tween fluoxetine (or any antidepressant) and suicidal behaviors. However, as with any rare event, one cannot totally discount the possibility that antidepressants may “trigger” suicidal ideation, either through a direct neurochemical effect (e.g., through an acute decrease in serotonergic transmission) or through nonspe-cific side effects of the drug (e.g., the induction of akathisia). If such a phenomenon exists, it is very uncommon, and probably not specific to any one antidepressant. There is, however, now a “black box” warning regarding the emergence of suicidal idea-tion in some adolescents taking SSRIs.
These concerns must be balanced over the well-proven and much greater risk of suicidal ideation and attempts when depression is not treated. Given this, and the propensity of suicidal patients to choose medication overdose as the method of suicide, the wide safety margin of the second- and third-generation antidepressants makes them preferable to first-generation agents when suicide is a concern. Bupropion has generally not been fatal in overdose, although about a third of such cases experience a seizure.