Drug-induced falls in older persons: is there a role for therapeutic drug monitoring?

Abstract

Background:

Falls are the leading cause of injuries among older persons. Because of ageing societies worldwide, falls are expected to become a prominent public health problem. The usage of several types of drugs has been associated with an increased fall and fracture risk. In order to reduce future falls, preventative measures are needed. Therapeutic drug monitoring may help to identify persons who are at risk for falls due to drug use. The aim was to demonstrate how drugs can contribute to falls and the role of therapeutic drug monitoring.

Methods:

We present a descriptive case series of four patients.

Results:

All patients were referred to the geriatric outpatient clinic of a university medical center. The presented cases contained different underlying mechanisms contributing to an increased fall risk in older adults, including renal failure, genetic variation, overdose and ageing.

Conclusion/discussion:

Older adults are more prone to the side effects of drug use, including falls. Therapeutic drug monitoring may be useful to identify the patients who have an increased drug-related fall risk and to prevent future falls by individualizing the drug regime.

Introduction

Falls in older persons are associated with a high morbidity and mortality [Kannus et al. 1999; Stevens et al. 2006] and result in high numbers of emergency department visits and hospital admissions [Hartholt et al. 2011; Boye et al. 2014]; in the United States, 559,000 people aged ⩾65 years are admitted due to a fall annually [Hartholt et al. 2011]. Because of ageing societies worldwide, falls are expected to become a more prominent public health problem. Over the past decades, more than 20 risk factors for falls have been identified including drug use [Tinetti et al. 1988; Woolcott et al. 2009]. A number of commonly used drugs have been associated with an increased fall and fracture risk [Leipzig et al. 1999a, 1999b; Woolcott et al. 2009; Verdel et al. 2010]. The use of sedatives, hypnotics, antidepressants and benzodiazepines has been established as a significant risk factor for falls in elderly individuals [Woolcott et al. 2009]. Other drug classes associated with an increased fall risk are antihypertensives, diuretics, nonsteroidal anti-inflammatory drug (NSAIDs) and β-blockers [Leipzig et al. 1999a, 1999b; Woolcott et al. 2009].

The number of falls caused by drugs and/or drug intoxications is not known because falls are not officially recognized as an adverse drug reaction. A decrease in fall risk has been demonstrated after the reduction or cessation of fall risk increasing drugs [Campbell et al. 1999; van der Velde, 2005; Iyer et al. 2008]. The mechanisms by which drugs contribute to fall risk, as well as the role of therapeutic drug monitoring (TDM) are not well-established. In this paper, four cases will be discussed in terms of drug-induced falls and the possible role of TDM.

Patient A

A 57-year-old woman presented with frequent falls for 2 years. Because of apathy after a cerebrovascular accident, the psychiatrist had started her on amantadine, without effect. Furthermore, she was on dipyridamol twice daily (b.i.d.) 200 mg, candesartan once daily (o.d.) 16 mg, lisinopril o.d. 20 mg, carbasalate calcium o.d. 100 mg, amantadine three times a day (t.i.d.) 100 mg and simvastatin o.d. 40 mg. On examination she had orthostatic hypotension; blood pressure (BP) supine of 110/70 mmHg, 97/56 mmHg on 3 minutes standing, and a regular heart rate of 74/min. Serum sodium was 142 mmol/l, potassium 4.9 mmol/l, urea 12.2 mmol/l, creatinine 159 μmol/l and the estimated glomerular filtration rate (eGFR) 29 ml/min. Because of the impaired renal function, amantadine was stopped immediately after her first visit to the geriatric outpatient clinic (OPC). A serum level of amantadine was in the toxic range, 2.8 mg/l (toxic >1.0 mg/l). After discontinuing the amantadine, her orthostatic hypotension disappeared and she had no further falls (follow up 2 years).

Patient B

A 74-year-old female presented with numerous falls occurring after rising from chair or bed. The general practitioner had started betahistine t.i.d. 16 mg, without effect. In addition, she was on bromazepam t.i.d. 3 mg and citalopram o.d. 30 mg. On examination, her BP was 140/80 mmHg, with no orthostatic hypotension. The serum citalopram level was in the therapeutic range and the bromazepam level was in the toxic range 0.55 mg/l (therapeutic 0.05–0.2 mg/l; toxic level >0.3 mg/l). Bromazepam was reduced to t.i.d. 1.5 mg; at 3 months follow up she had had no further falls. However, because she felt unwell, she resumed the original dosage of bromazepam and started to fall once more.

Patient C

A 70-year old male was referred to the geriatric OPC because of 1–2 falls a week. He had a medical history of alcohol abuse and behavioural problems, and had become psychotic during diazepam withdrawal in the past. His psychiatrist had achieved a stable situation with the current drug regimen. Patient C used diazepam t.i.d. 5 mg, olanzapine o.d. 7.5 mg and venlafaxine o.d. 150 mg and o.d. 75 mg. On examination, he was depressed and nervous with a BP of 145/85 mmHg, and no orthostatic hypotension. γ-Glutamyl transpeptidase (GGT) was 70 units (U)/l and alanine aminotransferase (ALAT) 39 U/l. Because olanzapine, diazepam and venlafaxine can contribute to falls and dizziness, the serum level of all three drugs was obtained. Olanzapine was in the therapeutic range. The level of desmethyldiazepam, the active metabolite of diazepam, was 1702 μg/l (therapeutic range 200–600 μg/l). Levels of venlafaxine and the active metabolite O-desmethylvenlafaxine were 1180 μg/l (therapeutic range 200–600 μg/l, toxic >1000 μg/l) and 29 μg/l (therapeutic 150–500 μg/l), respectively. Because of his previous psychotic episode on diazepam withdrawal, the psychiatrist preferred not to reduce the diazepam and venlafaxine dosages, despite the problem of falls in this patient.

Patient D

An 80-year-old woman was referred to the geriatric OPC because of falls twice a month. She used nortriptyline o.d. 50 mg, nifedipine o.d. 30 mg, labetalol b.i.d. 200 mg, bromazepam o.d. 1.5 mg and omeprazol o.d. 20 mg. She had orthostatic hypotension, BP 153/70 mmHg supine and 130/60 mmHg after 3 minutes standing. Her serum level of nortriptyline was 196 μg/l (therapeutic range 75–250 μg/l). Because nortriptyline can induce orthostatic hypotension, the dosage was reduced to o.d. 25 mg. At follow up 3 months later, she had not had any further falls and the serum level of nortriptylin had fallen to 65 μg/l.

Discussion

The four patients presented in this case series all suffered from drug-induced falls. It was already known that the use of certain drugs is associated with an increased fall risk. As our four cases illustrate, paying attention to the patient’s drugs as a potential cause of the falls can reverse the situation. Older patients are more susceptible to adverse drug reactions such as falls than younger patients [Bressler and Bahl, 2003] due to increased drug use, reduced physiological functions, and altered pharmacokinetics and pharmacodynamics [van Harten, 1995].

Multiple underlying causes can contribute to an interindividual pharmacological response, such as renal or liver impairment, or genetic causes, i.e. cytochrome P450 (CYP450) polymorphisms. Therefore, not only the number, but also the type of drugs, and possible interactions are important. TDM is the regular measurement of serum drug levels to titrate drug doses. Therapeutic and toxic drug plasma concentrations are available for a wide variety of drugs. However, these concentrations were assessed in nonelderly patients, so it is questionable whether they apply to older patients. Nevertheless, TDM in older patients is advisable for drugs with a narrow therapeutic window, such as anticonvulsants, antidepressants and antipsychotics.

Patient A was treated with amantadine 100 mg t.i.d. Amantadine is primarily excreted unchanged in the urine by glomerular filtration and tubular secretion. In patients with a GFR of 29 ml/min, a dose of 100 mg once every three days is recommended. In patient A, the dose of 100 mg t.i.d. resulted in a toxic plasma level of 2.8 mg/l, resulting in orthostatic hypotension and falls.

Patient B used bromazepam and could not discontinue this drug due to her panic disorder. Bromazepam is metabolized to 3-hydroxybromazepam by CYP450 1A2 [Sachse et al. 1997]. Benzodiazepines are widely prescribed among the elderly and increases the risk of falls with 60% [Andersson et al. 1994]. A daily dose of 9 mg is the maximum dose, although in severe cases higher doses can be prescribed. In older patients, a dose reduction is recommended. In this patient, the plasma level of bromazepam was in the toxic range and was the result of a decreased hepatic elimination. Since there was no indication of liver dysfunction, the decreased hepatic elimination may be the result of aging.

Patient C received subtherapeutic to therapeutic doses of three psychotropic drugs: diazepam, olanzapine and venlafaxine. The plasma levels of venlafaxine and desmethyldiazepam, an active metabolite of diazepam, were respectively two and three times the upper limit of the therapeutic range. Venlafaxine is metabolized to the active metabolite O-desmethylvenlafaxine by the CYP2D6 enzyme, and O-desmethylvenlafaxine is conjugated to an inactive metabolite. Both metabolites are excreted renally. In this patient a high venlafaxine level and a low O-desmethylvenlafaxine level was measured, suggesting a poor functioning of the CYP2D6 enzyme. The patient did not use medication that inhibited the CYP2D6 enzyme and therefore we suspected a genotype coding for a poorly functioning CYP2D6 enzyme. The patient’s genotype was indeed CYP2D6 *4/*4 coding for two nonfunctioning CYP2D6 enzymes. In a Caucasian population, approximately 5–10% of the population has no functioning CYP2D6 enzymes [Turnheim, 1998]. Diazepam undergoes demethylation by the CYP2C19 enzyme to desmethyldiazepam and hydroxylation by the CYP3A4 enzyme to temazepam and oxazepam. The levels of diazepam were in the therapeutic range, while the levels of desmethyldiazepam were above the therapeutic range. This may indicate that, in patient C, the functioning of the CYP2C19 enzyme was much better than the functioning of the CYP3A4 enzyme. Genetic variation resulting in a poor functioning CYP3A4 enzyme is rare. With aging and decreased hepatic functioning, the cytochrome enzymes are affected differently. It has been suggested that the CYP3A4 enzymes are more prone than the CYP2C19 enzymes, which is in line with the higher desmethyldiazepam levels [Sachse et al. 1997]. The fragile balance between inappropriate behavior and psychosis on diazepam withdrawal in this case demonstrates the difficulty of pharmacotherapy in some patients and that some side effects should be accepted in order to prevent a new psychosis.

Patient D was using a therapeutic dosage of nortriptyline and her plasma levels were within the therapeutic range. Nonetheless, she experienced falls induced by nortriptyline (orthostatic hypotension as an anticholinergic side effect). Reducing the dosage stopped the falls and the orthostatic hypotension. This case illustrates that plasma levels within the therapeutic range do not exclude drug-induced falls.

Conclusion

The four patients in this case series all presented with drug-induced falls, three of whom (A, B, C) had a drug intoxication though the underlying mechanism varied. Patient A had decreased renal functioning and was prescribed a renally excreted drug without dose adjustment. Patients B and C had decreased hepatic metabolism, likely related to aging; in addition, patient C had an inherited absence of the CYP2D6 enzyme. The prescribed drug regimens resulted in toxic plasma levels in these patients. Patient D had serum levels within the therapeutic range, but had falls due to the anticholinergic side effects of nortriptyline. Lowering the dose stopped the falls. Most likely, this patient had a higher sensitivity due to pharmacodynamic changes. Drug-related falls are potentially preventable and TDM can help in identifying patients with drug-induced falls and confirm clinical suspicion. However, plasma levels within the therapeutic range do not exclude the diagnosis of drug-induced falls.