Selective serotonin-reuptake inhibitors

With the exception of paroxetine which may have a modest effect on cholinergic
receptors (Thomas et al. 1987), the SSRIs have little affinity for adrenergic, histaminergic, or
cholinergic receptors (Grimsley and Jann 1992). In selected physically healthy individuals,
SSRIs lack significant cardiac effects and do not produce postural hypotension (Grimsley and
Jann 1992). However, their safety in patients with cardiac disease has not been well studied
(Sheline et al. 1977). A recent controlled trial found that paroxetine and nortriptyline had
comparable efficacy in the treatment of major depression in patients with ischemic heart disease,
but paroxetine was associated with significantly fewer adverse cardiac events (Roose et al.
1998). Consistent with their apparent lack of cardiotoxicity, SSRIs have not proved fatal when
taken alone in overdose. In contrast to TCAs, SSRIs have not been found to impair cognitive
function (Hindmarch et al. 1990, 1994). Patients with dementia, therefore, may better tolerate
SSRIs, although a direct comparison of an SSRI with approximately selected TCA (that is, a
secondary amine drug) has not been undertaken in the patients. Thus, SSRIs lack many of the
adverse effects associated with TCAs and, as a result, there are fewer barriers to the use in the
elderly. However, SSRIs are not without adverse effects; although one may surmise that they are
better tolerated than second amine TCAs by medically ill older patients this hypothesis has not
yet been tested in a randomized controlled trial.
The adverse effects of SSRIs primially reflect central and peripheral serotonic-reuptake
inhibition, including gastrointestinal upset (nausea, anorexia, flatus, frequent or loose stools),
CNS dysfunction (insomnia, sedation, apathy, nervousness, restlessness, dizziness, tremor,
headaches), and sexsual dysfunction (decreased libido, anorgasmia, delayed ejaculation, and
impotences) (Grimsley and Jann 1992)
A fes side-effects appear to be more prevalent or disabling in older persons. SSRIs have
been found to produce postural instability in older person (Laghrissi Tode et al. 1995) and, like
many other psychotropic drugs, have been associated with an increased risk of falls and hip
fractures in the elderly (Liu et al. 1998). Tremor may also be more problematic in older persons,
especially if an SSRI is combined with lithium or neuroleptic. However, open-label trials in
patients with treated Parkinsons disease have not identified a worsening of motor symptoms
with SSRI use (Meara et al. 1996; Hauser and Zesiewicz 1997). An infrequent, but potentially
dangerous, adverse effect is the inappropriate secretion of antidiuretic hormone (SIADH) and

hyponatraemia (Liu et al. 1996) which can present with confusion, fatigue, weakness, and falls.
Most reported cases of SSRI-induced hyponatraemia have occurred within the first month of
starting the drug (Liu et al. 1996). The hypronatraemia will resolve once the SSRI is
discontinued (a short period of fluid restriction may also be necessary to correctthe electrolyte
imbalance), but may reoccur f the patient is rechallenged with the same or different SSRI (Flint
et al. 1996). Although SIADH has been associated with other classes on antidepressants, it has
been reported most frequently with SSRIs. However, it is not known this reflects a reporting bias
or an increased incidence of this effect with SSRIs.
There are differences in the pharmacokinetics of SSRIs that have implication for their
use in the elderly. Fluoxetine and its active metabolite norfluoxetine have a combined
elimination half-life of approximately 2 weeks (Rickels and Schweizer 1990). As a result, it may
takes 2 moths or more to achieve steady-state plasma concentrations of this drug, and there is
potential for ongoing side-effects and drug-drug interactions for many week after the drug is
discontinued. Because of its long half-life, fluoxetine is potentially more problematic than
citalopram, fluvoxamine, paroxetine, or sertraline, which have half-lives of 1-2 days and
metabolites with little or no clinically significant activity (Grimsley and Jann 1992; Goodnick
1994). Fluoxetine and paroxetine have non-linear kinetics, and thus an increase in their dose can
result in a disproportionate increase in plasma concentrations (Goodnick 1994). Therefore, the
dosage of fluoxetine or paroxetine should not be increased by any more than 10 mg at a time in
older patients. Age-related pharmacokinetic changes result in higher plasma concentrations and
prolonged half-lives of citalopram and paroxetine, hence the starting dose of these drugs should
be reduced in the elderly (Goodnick 1994) (Table 122). Age, per se, has not been shown to have
a clinically significant effect on the pharmacokinetics of sertraline of fluvoxamine. Nevertheless,
medically ill older persons may better tolerate these medications if they are started at a low dose
and then titrated up to the minimum effective dose after a week or so (Table 12.2).
SSRIs also differ in their potency of inhibition of CYP isoenzymes (Nemeroff et al. 1996;
Shader et al. 996) (Table 12.3) it is important to note that the clinical significance of many
potential SSRI-drug interactions is not known. Even when they are significant, there is
tremendous interindividual variability in the severity of the clinical sequelae. Therefore, a
potential drug-drug interaction dose may need to be adjusted and the patient should be carefully
monitored. Particular caution should be exercised when an SSRI is prescribed with a drug that
gas a narrow therapeutic index such as TCAs, antipsychotics, theophylline, phenytoin,

carbamazepine, and tolbutamide (Nemeroff et al. 1996). SSRIs should be administered with
caution, if at all, in patients taking type 1C antiarrhythmics (encainide, flecainide, propafenone)
(Nemeroff et al. 1996).

Table 12.2 usual starting dose, minimum effective dose, and maximum dose of selective serotinreuptake inhibitors in elderly patients
SSRI
Citalopram
Fluoxetine
Fluvoxamine
Paroxetine
Sertraline

Usual starting
Dose (mg/d)
10
10
25-50
10
25-50

Minimum effective
Dose (mg/d)
20
20
100-150
20
50-100

Maximum dose
(mg/d)
30
60
300
40
200

Table 12.3 Inhibitory effect of selective serotonin reuptake inhibitors on P450(YCP)

isoenzymes*
SSRI
Isoenzyme 1A2
2C
2D6
3A4
Citalopram
0
0
+
0
Fluoxetine/norfluoxetin 0
++
+++
++
e
+++
++
0
++
Fluvoxamine
0
0
+++
0
Paroxetine
0
+
+/++
+
Sertraline
Potency of inhibition 0 = none or nominal, + = mild, ++ = moderate, +++ = potent
A rating of ++ or higher is usually necessary for inhibition to have clinical significance
*Based on currently available in vitro and in vivo data (Crewe et al. 1992; Goodnick 1994;
Harvey and Preskorn 1996; Nemeroff et al. 1996; Shader et al. 1996; Sproule et al. 1997)
SSRIs have a flat dose-response curve. For the majority of patients, increasing the dose
beyond the minimum effective dose does not result in additional efficacy, but can produce more
side-effects (Preskorn 1993). Thus, a patient should be monitored on the minimum effective dose
for at least 4 weeks before an increase in dose is considered. If there are no signs of clinical
improvement, the dose should then be gradually increased, guided by tolerability and efficacy.
Other antidepressants
Mirtazapine, nefazodone, and venlafaxine have minimal anticholinergic effects (Holm
and Markham 1999; Robinson et al. 1996; Feighner 1994, respectively). in addition, they have
not been found to have clinically significant effects on the heart., although there are few data

pertaining to their use in patients with cardiovascular disease (Feighner 1994); Robinson et al.
1996; Holm and Markham 1999). Similar to SSRIs, they do not appear to be lethal when take
alone in overdose (Feighner 1994 Robinson et al. 1996; Holm and Markham 1999). Mirtazapine
and venlafaxine do not cause orthostatic hypotension (Feighner 1994; Holm and Markham
1999).
Nefazodone is structurally related to trazodone. Like trazodone, it is a weak inhibitor of
serotonin reuptake but a potent serotonin 5HT2 receptor antagonist (Goodnick 1994). Although
nefazadone has affinity for @ adrenergic receptor, it is approximately five times less potent than
trazodone in this regard (Taylor et al. 1995) and arthostatic hypotension is nor a common
complication (Goldberg 1997). Nefazodone can cause daytime sedation, but this effect is less
marked than with trazodone (Robinson et al. 1996). Other side-effects include nausea, dry
mouth, dizziness, and asthenia (Robinson et al. 1996). In contrast to SSRIs, nefazodone does not
caus sexual dysfunction and unlike trazodone ti does not cause priapism (Robinson et al. 1996).
Nefazodone is a potent inhibitor of CYP3A4 (Nemeroff et al. 1996) and care should be taken
when it is administered with other drugs metabolized by this isoenzyme. Ageing affects its
pharmacokinetic, with the result that older patients require half the usual starting dose (50 mg
twice a day), slower dose titration, and a lower therapeutic dose (200-400 mg/day) (Barbhaiya et
al. 1996).
Venlafaxine inhibits serotonin and noradreanile reuptake and, in this respect, is
pharmacodynamically similar to TCAs (Feighner 1994). However, venlafaxine lacks TCAs
unwanted affinity for cholinergic, histamirgenic, and @ adrenergic reseptors (Feighner 1994).
The most common adverse effects of venlafaxine are nausea, somnolence, dry mouth, dizziness,
nervousness, constipation, sexual dysfunction, sweating, and asthenia (Feighner 1994). A small
number of patients experience modest elevation of blood pressure at the higher doses of this
medication and, as a result, blood pressure monitoring is recommended before and during
venlafaxine therapy (1994). However, as long as a patients blood pressure is adequalety
controlled, a history of hypertension does not contraindicate the use of this drug. Venlafaxine is
only a weak inhibitor of CYP2D6 and is much less likely than fluoxetine to interact with drugs
metabolized by this isoenzyme (Nameroff et al. 1996). Unlike SSRIs, venlafaxine does not have
a flat dose-response curve, and higher doses of this drug are generally associated with a higher
probability of response (Feighner 1994). However, the dose range is wide established.
Furthermore, it has not been determined whether there is therapeutic range of plasma

concentration for this drug. In older persons, a starting dose of 37.5 mg/day is recommended in
order to avoid nausea.
Mirtazapine enhances noradrenergic and serotonergic transmission via blockade of @2,
5HT2, and 5HT3 receptors (Holm and Markham 1999). Mirtazapine has a high affinity for
histamine H1 receptors and, as a result, can cause sedation, increased appetite, and body weight
gain (Holm and Markham 1999). Interestingly, sedation may diminish at higher doses, possibly
because of increased arousal associated with @2 antagonism (Holm and Markham 1999). In
contrast with other newer antidepressants, mirtazapine does not cause nausea, presumably
because of 5HT3 blockade (Holm and Markham 1999). Limited data suggest that mirtazapine
has little impact on sexual function (Holm and Markham 1999). In vitro and in vivo studies
indicate that mirtazapine is unlikely to affect the metabolism of drugs metabolized by CYPIA2,
CYP2D6, and CYP3A4 (data are not available regarding CYP2C) (Holm and Markham 1999).
Ageing does not appear to have a clinically significant effect on the pharmacokinetics of this
drug, and the recommended starting dose (15mg/day) and effective dose range (15-45 mg/day)
are the same as those for younger adults (Holm and Markham 1999)
Stimulants
Uncontrolled data and the result of one small randomized, controlled, crossover study
suggest that stimulant such as methylphenidate may be useful in the treatment of depression or
apathy in medically ill patients (Satel and Nelson 1989; Wallace et al.1995; Emptage and Semla
1996). At the recommended doses (for example, 10-30 mg/day of methylphenidate) stimulant are
generally well tolerated, with a rapid onset of action (often within days) reported (Satel and
Nelson 1989; Emptage and Semla 1996). However, there are virtually no data on the mediumterm or long-term efficacy of these drugs, in particular their ability to prevent relapse or
recurrence of depression (Emptage and Semla 1996). Furthermore, there are no randomized
controlled data on whether stimulants are as effective as conventional antidepressants in the
treatment of major depression in medically ill older patients (Emptage and Semla 1996).
However, there are considerable data from controlled trials to show that stimulants are usually no
more effective that placebo in the treatment of primary depression in people who are physically
well (Satel and Nelson 1989). Therefore, in the opinion of this author, conventional
antidepressants, rather than stimulants, are the drugs of choice of the treatment of major
depression, regardless of whether physical illness is also present. Uncontrolled data suggest data

methyphenidate may be useful in the treatment of apathy that is secondary dementia (Galynker et
al. 1997)
Treatment of non-responders
The goal of acute phase of treatment is to achieve the remission of depressive symptoms.
The first step is to ensure that the patient has received an adequate dose of antidepressant for a
sufficient length of time (usually a minimum of 6 weeks). However, 30% of elderly patient do
not respond to an initial adequate trial of antidepressant medication and require additional or
alternative treatment (Schneider 1996). One approach is to augment the antidepressant with
another drug, either a second antidepressant (for example, adding lithium, tri-iodothyronine,
methylphenidate, pindolol, buspirone, or valproate) (Flint 1995). The advantage of augmentation
is that it does not require discontinuation of the original antidepressant and, therefore, patients
who partially respond to treatment are not put at risk of returning to their baseline severity of
depression. Also, response may at times be faster with augmentation than with a new trial of antidepressant medication. The disadvantage of augmentation, especially in medically ill elderly
patients, is that the combination of medications increases the risk of adverse effects and drugdrug interactions. Furthermore, there have been no controlled trials of augmentation in elderly
people and, therefore, the effectiveness of this approach in treating geriatric depression has not
been established (Flint 1995).
Refractory depression can also be managed by changing from one antidepressant
medication to another. Generally, patient should be given a different class of anti-depressant
rather than drug in the same class (Thase et al. 1994/1995). If a patient has failed to respond to
an SSRI or nefazadone, then changing to anti-depressant with dual neurotransmitter action (for
example, venlafaxine, mirtazapine, or notriptyline) is a sensible strategy. Finally,
electroconvulsive therapy (ECT) can be an efficacious and safe treatment for depression in
medically ill older patients and should always be considered as an alternative to
pharmacotherapy in this population (Greenberg 1997).
Anxiety
When anxiety develops in elderly patients with medical illness or neurological disorder it
is frequently associated with depression (Flint 1994). Therefore, new-onset anxiety, especially if
the symptoms are of generalized anxiety or panic, should always prompt a careful examination
for an underlying depressive illness. If anxiety I found to be associated with depression, the most
appropriate primary pharmacological treatment is antidepressant medication (Flint 1998b).
Certain antidepressants can also be used as a first-line treatment for generalized anxiety disorder,

panic disorder, obsessive-compulsive disorder, post-traumatic stress disorder and social phobia,
even when depression is absent (Antony and Swinson 1996). SSRIs in particular have broad
spectrum of efficacy in the treatment of primary or secondary anxiety (Antony and Swinson
1996).
Benzodiazepines may at times be required as an adjunct to antidepressant medication in
treatment of anxiety. In addition, benzodiazepines may be used as a primary treatment for acute
situational anxiety, generalized anxiety disorder or panic disorder that are not associated with the
depression, and social phobia (Antony and Swinson 1996). The benzodiazepines that are most
appropriate for use in the elderly are lorazepan or oxazepan since they have relatively short halflives (12-15 hours), no active metabolites, and do not undergo oxidative metabolism in the liver
(and thus their clearance is unaffected by age or other drugs). As a result, these drugs do not have
the potential for cumulative toxicity, which is a risk associated with longer acting agents such as
diazepam (American psychiatric Association 1990). The usual dose range in the elderly is 0.5-2
mg/day of lorazepam and 10-30 mg/day of oxazepam.
Although buspirone can be an effective treatment for generalized anxiety disorder
(Steinberg 1994), it has a limited role in the management of anxiety in later life. Buspirone has a
delayed onset of action of 2 weeks or more (Steinberg 1994) and so it is unsuitable for acute
anxiety states. Furthermore, as already noted, generalized anxiety in the elderly is frequently
associated with a depressive illness and, in this situation, antidepressant medication remains the
drug of choice. Unlike benzodiazepines, buspirone does not cause sedation, psychomotor or
cognitive impairment, or depressed respiratory drive (Steinberg 1994). Therefore, buspirone may
be preferable to a benzodiazepine in the longer term management of generalized anxiety in nondepressed patients with chronic obstructive airways disease, sleep apnoea, or neurological
disorders.
Bipolar disorder (mania)
Lithium has been the mainstay of the acute and maintenance treatment of bipolar disorder
in late life (Eastham et al. 1998). However, there have been no controlled studies of its use in the
elderly. Because of an age-related decline in renal clearance, older patients require lower doses
of lithium to achieve therapeutic plasma levels (Eastham et al. 1998). The elderly are particularly
sensitive to the neurological effects of lithium (tremor, ataxia, cognitive impairment) even at
plasma concentrations considered to be therapeutic in younger adults (Flints 1993). This
vulnerability in enhanced when lithium is combined with other medications that have

neurological effects (for example, neuroleptics or anticonvulsants) or if the patient has

neurological disease. As a result, plasma levels of 0.5-0.8 mmol/l have been recommended for
use in the elderly (frequently achieved with doses in the range of 300 to 750 mg/day) (Eastham
et al. 1998). Thiazide diuretics, angiotensin-converting enzyme (ACE) inhibitors, and nonsterodial anti-inflamantory drugs can interfere with the excretion of lithium, and caution should
be exercised when these drugs are used in combination with lithium.
In recent years, valproate and carbamazepine have been widely used as alternatives or
adjuncts to lithium in the treatment of mania and bipolar disorder (Baldssarini et al. 1996). These
drugs have not been systematically studies in the elderly, but in younger individuals they have
been shown to be effective therapies for mania, including mania that has failed to respond to
lithium (Baldessarini et al. 1996). Controlled studies in mixed-age patients have found that
carbamazepine is as effective as lithium in preventing recurrences of bipolar disorder, and
uncontrolled trials have provided encouraging observations about the prophylactic effect of
valproate (Baldessarini et al. 1996). Uncontrolled data suggest that valproate and carbamazepine
may be more efficacious than lithium in the treatment of rapid cycling bipolar illness
(Baldessarini et al. 1996). As previously noted, anecdotal evidence suggests that valproate may
be better tolerated as carbamazepine in old age. The recommended plasma level for valproate in
the treatment of bipolar disorder in the elderly is 50-100 ug/ml (350-700umol/l). Combination
mood-stabilizer therapy (for example, lithium and valproate, or lithium and carbamazepine) is
commonplace in the treatment of refractory bipolar disorder in younger individuals (Freeman
and Stoll 1998). Although some elderly patients may tolerate combination therapy, adverse
effects may limit this strategy in later life (Wils and Goluke-Willemse 1997).
Adjunctive treatment with antipsychotic medication is frequently required in acute mania,
especially is psychosis or severe behavioural disturbance is present. There have been a number of
reports of neurotoxicity resulting from the combination of lithium and antipsychotic medication
(Freeman and Stoll 1998). Most of these cases of neurotoxicity were associated with the use of
conventional antipsychotics (especially haloperidol), but the combination of lithium and
risperidone can also carry some risk (Freeman and Stoll 1998). Therefore, care should be taken
when combining lithium and a neuroleptic in the elderly, even if the antipsychotic is an atypical
agent. Neurological side-effects may possibly be less of a risk with a valproate-antipsychotic
combination (Freeman and Stoll 1998), but, nevertheless, patients should be closely monitored
for adverse effects. Antipsychotic medication is seldom indicated in the maintenance treatment of

geriatric bipolar disorder, and every effort should be made to withdraw an antipsychotic
following resolution of the acute manic episode.
Although efficacy data are lacking, benzodiazapines are occasionally administrated to
elderly patients with acute mania. Low doses of lorazepam may be used to treat sleep disturbance
or mild agitation when a mood stabilizer alone does not adequately control these symptoms.
Clonazepam is frequently used as an adjunctive agent in the acute treatment of mania in younger
patients. However, clonazepam is generally not recommended for older patients because it has a
long elimination half-life and the potential for accumulation and toxicity. At dosages as low as I
mg/day, clonazepam can cause clinically significant drowsiness and psychomotor impairment in
the elderly, especially when it is combined with antipsychotics, lithium, or anticonvulsants.
Schizophrenia, delusional disorder, and the other psychoses
Schizophrenia and delusional disorder
Irrespective of whether they are of early or late onset, schizophrenia and delusional
disorder are typically chronic disorder requiring long-term treatment. Elderly patients are at high
risk of developing tardive dyskinesia (Kane et al. 198) and this is an important consideration in
the selection of an antipsychotic I these disorders. Preliminary data suggest that older patients
may be less likely to develop tardive dyskinesia with risperidone than with haloperidol (Jaste et
al. 1999). Moreover, study involving younger patients with schizophrenia found that the
incidence of tardive dyskinesia was significantly lower with olanzapine than with haloperidol
(Tollefson et al. 1997). Thus, in this respect, atypical antipsychotics may have a major advantage
over conventional antipsychotics in the treatment of schizophrenia or delusional disorder in later
life.
Psychotic depression
Studies involving young and middle-aged adults have found that depression with
psychotic features (psychotic depression) is more likely to respond to a combination of tricyclic
antidepressant and antipsychotic medications than to a tricyclic alone (Kroessler 1985). In some
studies the rate of response to combination pharmacotherapy was similar to that associated with
ECT (Kroessler 1985); Spiker et al. 1985). In contrast, preliminary data suggest that a trycyclicantipsychotic combination has limited efficacy in the treatment of psychotic depression in older
patients (Meyer et al. 1985; Flint and Rifat 1998), and many psychiatrists consider ECT to be the
treatment of choice for this condition in late life (Finaly-Jones and Parker 1993). Currently, there

are no data on the efficacy or safety of newer antidepressant or antipsychotics in the treatment of
psychotic depression in the elderly.
Psychosis associated with Parkinsons disease
Hallucinations and delusions can occur in the patient with Parkinsons disease, usually in
association with the drugs used to treat this disorder. If psychotic symptoms are troublesome, the
first strategy is to attempt to reduce the dose of the antiparkinsonian dug or to use a different
drug. Patients with Parkinsons disease are particularly sensitive to the extrapyramidal effects
data are limited. A recently published, randomized, double-blind trial found that low doses of
clozapine (starting dose of 6.25-50 mg/day; dose range of 6.25-50 mg/day) was significantly
more effective than placebo in the treatment of psychosis in patients with Parkinsons disease
(The Parkinson Study Group 1999). At these doses, clozapine was generally well tolerated and
had no deleterious effects on the severity of Parkinsonism. In fact, interestingly, tremor improved
with clozapine. Case reports suggest that low doses of quetiapine (a starting dose of 12.5-25
mg/day is recommended) may also be effective and well tolerated in these patients (Parsa and
Bastani 1998).