FALLING AND OTHER ADVERSE EFFECTS

Although there are benefits for psychotropic medications for older people, significant adverse effects have been associated with their use, including diminished cognition, excessive sedation, and an increased risk of falls and fractures (Aparasu and Mort, 2004). This section of the report summarizes the literature on polypharmacy and falling in the older community-dwelling population, and concludes by detailing the effects of cardiac polypharmacy on visual performance.

In older people, benzodiazepines are principally used for insomnia and anxiety disorders. The acute use of benzodiazepines can lead to memory problems, perceptual errors, and impairments on tests of attention (Hogan, Maxwell, Fung, and Ebly, 2003). In a 5-year longitudinal study of subjects 65 and older, they found that benzodiazepines were associated with adverse outcomes. In particular, affect, self-rated health, cognition, function, and new institutionalization were significantly associated with benzodiazepine use. Benzodiazepine users had a lower ADL and IADL score (activities of daily living and instrumental activities of daily living), performed worse on a test of attention and psychomotor abilities (Digit Symbol Substitution Test), and had a higher likelihood of falling.

Tinetti (2004) states that each of the following conditions has been shown to increase the risk of falling in two or more observational studies:

• Arthritis
• Depressive symptoms
• Orthostasis
• Impairment in cognition, vision, balance, gait, or muscle strength
• Use of four or more prescriptions

Foster, Hillegass, and Phillips (2004) cite the Guidelines for the Prevention of Falls in Older Persons (2001), which includes a meta-analysis of studies examining both multiple and single risk factors for falling and medications, and found a significantly increased risk from psychotropic medications, class 1A anti-arrhythmic medications, digoxin, and diuretics. Tinetti (2004) reports that although there is a clear relation between falling and the use of a higher number of medications, the risks associated with individual classes of drugs have been more variable (Leipzig, Cumming, and Tinetti, 1999). To date, serotonin-reuptake inhibitors, tricyclic antidepressants, neuroleptic agents, benzodiazepines, anticonvulsants, and class 1A anti-arrhythmic medications have been shown to have the strongest link to an increased risk of falling (Leipzig et al., 1999; Thapa, Gideon, and Cost, 1998). A recent meta-analysis found that the odds ratio for falls was 1.48 with benzodiazepine use, with no difference between short- and long-acting preparations (Leipzig et al., 1999). In Hogan et al.’s (2003) sample of subjects 65 and older, those not currently taking benzodiazepines were less likely to have fallen, than those who were either continuous users or new users of benzodiazepines (33.5% versus 40.3%, p=0.48).

In a sample of community-dwelling subjects age 65 to 84 who participated in the Salisbury Eye Evaluation project, multiple medication use, defined as taking four or more medications, predicted fear of falling at a 20-month follow-up clinic evaluation for 1,466 of the 2,212 participants who did not express fear of falling at the baseline clinic evaluation (Friedman et al., 2002). In this study, fear of falling at baseline was a predictor of falling 20 months later. Of the 1,597 subjects who had not had a fall at baseline, sedative use and fear of falling at baseline were independent predictors of falls at the 20-month follow-up visit. Sedatives were defined as any benzodiazepines, phenothiazines, or antidepressants.

In a cross-sectional analysis of self-reported medication use and self-reported falling over the past 12 months by high-functioning community-dwelling black and white people age 70 to 79, fallers were more likely to report more medications than nonfallers (de Rekeniere et al., 2003). Fallers took a mean of 6.1 medications (sd=4.3) and nonfallers took a mean number of 5.5 medications (sd=3.9). This difference was significant at the 0.01 level. In women, benzodiazepine use was associated with past falls. In the multivariate model, the adjusted odds ratio for benzodiazepine use versus none in women was 1.6 (1.0-2.6). Participants included 3,075 people in the Health, Aging, and Body Composition Study identified from a random sample of white Medicare beneficiaries and all age-eligible black community residents in designated ZIP code areas surrounding the Pittsburg, PA, and Memphis, TN, study centers.

In a population of 489 patients admitted to the neurology department of a hospital in Kiel, Germany, during 100 consecutive days, Stolze et al. (2004) found that the average age of the 165 patients who had experienced a fall in the past 12-month period was significantly higher than the average age of the nonfallers (62.8 + 15.5 years versus 55.3 + 16.6 years, p< 0.001). Descriptive statistics showed clear differences in the percentage of fallers and nonfallers who were being treated with medications known to be risk factors for falls, as follows:

A logistic regression analysis showed a significant correlation between the occurrence of falls and medication for digitalis (r=0.064, p<0.05), diuretics (r=0.061, p<0.05), and neuroleptics (r=0.62, p< 0.05). The number of falls was correlated ( Kendall’s t-B) with antihypertensiva (t-B=0.15, p<0.05), benzodiazepines (t-B=0.12, p<.05), and neuroleptics (t-B=0.19, p<0.05). The authors conclude that medications that lower blood pressure have the potential to cause falls, irrespective of neurological diagnosis.

Kennedy et al. (2002) compared the accident types of patients admitted to trauma centers in Scotland, a percentage of whom were insulin-treated diabetics. Accidents in this study included motor vehicle crashes, assaults, falls greater than 2 meters (6.56 ft), falls of 2 meters (6.56 ft) or less, sports injuries, and other unclassified injuries. They found that insulin-treated diabetics had an accident rate (across all accidents) that was 291.2 per 100,000 population per year compared to an accident rate of 148.4 for the control population (this difference is significant, with a relative risk of 1.97). The insulin-treated patients were significantly older, more likely to be women and had a longer stay in the hospital, than control patients. The major injury type was low falls (less than 2 meters), accounting for 62 percent of the injuries to the insulin-treated group and 47 percent of the control group. Of the 151 injuries sustained by the 94 insulin-treated patients with low falls, the majority were fractures at peripheral sites (fractures of the femur, fibula, tibia, humerus, radius, ulna, calcaneus, and digit accounted for 75% of the injuries). There were only 23 motor vehicle crashes (MVCs) among the insulin-treated patients during the study period. The percentage of accidents resulting from MVCs was lower in the insulin-treated group than in the remainder of patients in the database (15% versus 24%).

Lawlor, Patel, and Ebrahim (2003) found that chronic diseases and multiple pathology are more important predictors of falling than polypharmacy. They used a cross-sectional survey design and data from the British women’s heart and health study for 4,050 women ages 60 to 79. Slightly over 70 percent of the women were taking at least one drug, and 15 percent were taking five or more drugs. Although there was a strong linear association between the number of drugs taken and whether they had a fall in the previous 12 months (crude odds ratio = 1.14 for each additional drug taken), the association was not significant when the data were adjusted for chronic diseases and other potential confounding factors (fully adjusted odds ratio = 1.01). Only two classes of drugs—hypnotics or anxiolytics, and antidepressants—were independently associated with increased odds of falling, even with adjustment for chronic disease status (including ever having a diagnosis of depression) and other potential confounding factors (e.g., age, body mass index, hemoglobin concentration, heavy alcohol consumption, and social class). Each class was associated with an increase of about 50 percent in the odds of falling. In the fully adjusted analyses, analgesics, cardiovascular system drugs, endocrine system drugs, and respiratory disease were not independently associated with having a fall. Nearly 75 percent of the women had at least one chronic disease. There was a significant linear trend of increasing odds of falling with increasing number of chronic diseases, even after adjustment for drug use and other potential confounding factors. The fully adjusted odds ratio for any fall in the previous 12 months associated with having at least one of the chronic diseases was 1.81, and for each additional simultaneously occurring disease was 1.37.

Lawlor et al. (2003) found that circulatory disease, chronic obstructive pulmonary disease, depression, and arthritis were each associated with higher odds of falling. The population-attributable risk of having had at least one fall in the previous 12 months (from the fully adjusted models) was 6.2 percent for coronary heart disease, 6.2 for circulatory disease, 8.0 for chronic obstructive pulmonary disease, 9.4 percent for depression, and 17.4 percent for arthritis. When number of drugs taken and number of chronic diseases were included in the same regression model, they combined multiplicatively. The odds ratio for a fall for each additional chronic disease, adjusted for number of drugs taken, was 1.39, and that for each additional drug taken, adjusted for number of chronic diseases was 1.05. There was no strong evidence of a statistical interaction between number of drugs and number of chronic diseases, and no evidence of statistical interactions between any of the individual chronic diseases and their relevant treatment. The population-attributable risk of falling associated with having any chronic disease was 32 percent, compared to that associated with the use of psychotropic drugs (between 2% and 5%).

The South Western Sydney Area Health Service Falls Subcommittee has listed the following classes of drugs associated with falls:³⁷

• Anxiolytics or sedatives (alprazolam, bromazepam, clonazepam, diazepam, lorazepam, oxazepam, flunitrazepam, nitrazepam, temazepam)
• Antipsychotics (chlorpromazine, fluphenazine, trifluoperazine, thioridazine, haloperidol, clozapine)
• Opioids (morphine, codeine, oxycodone, pethidine, methadone)
• Antidepressants (amitriptyline, nortriptyline, clomipramine, desipramine, doxepin, dothiepin, imipramine, mianserin, phenelzine, tranylcypromine, moclobemide, fluvoxamine, fluoxetine, paroxetine, citalopram)
• Antiparkinsonians (levodopa, bromocriptine, selegiline, benzhexol, benztropine, biperiden, orphenadrine, procyclidine)
• Hypoglycaemics (insulin, glibenclamide, gliclazide, glipizide, tolbutamide, metformin)
• Antihypertensives (captopril, enalapril, fosinopril, lisinopril, perindopril, quinapril, remipril, trandolapril, candesartan, irbesartan, losartan, telmisartan, amlodipine, dilatizem, nifedipine, felodipine, verapamil, atenolol, pindolol, propanolol, metoprolol, labetolol, prazosin, clonidine)
• Diuretics (frusemide/furosemide, ethacrynine/ethacrynic acid, buemtanide/bumetanide, bendrofluazide, chlorothiazide, hydrochlorothiazide, methyclothiazide)
• Laxatives/stool softeners within previous 24 hours

In consideration of “other adverse effects” of using multiple medications, cardiac polypharmacy has been associated with color vision and acuity changes in older people (Castells, Teitelbaum, and Tresley, 2002). In particular, the addition of amiodarone therapy in patients using digoxin has been associated visual “shining,” glare, color vision anomalies, and decreased visual acuity. Amiodarone is a benzofuran derivative, and is an approved anti-arrhythmic agent indicated for the treatment of ventricular tachyarrhythmias and other cardiac conditions. While both drugs can cause permanent visual changes, the ocular effects are often reversible, indicating the importance of ocular examination at baseline and every 6 months for patients receiving a regimen of amiodarone and/or digoxin therapy. Digoxin (Lanoxin) is an approved treatment for congestive heart failure and certain cardiac arrhythmias. Its ocular side effects overlap with those of amiodarone, and include color vision deficits (commonly yellow-brown), and central or paracentral visual-field defects. Unique to digoxin is a reported side effect of a white or colored “snowy” or “frosted” appearance to vision. Digoxin has a low safety profile because the therapeutic and toxic serum levels are known to be close. Digoxin toxicity is a result of both dosing and cumulative effects. It is estimated that up to 20 percent of patients who receive digoxin demonstrate some level of toxicity. Up to 95 percent of patients with toxic levels of digoxin demonstrate visual symptoms (Piltz et al., 1993).

A number of drugs can significantly increase serum levels of digoxin, easily inducing toxic levels. These include amiodarone, furosemide, nifedipine, verapamil, and quinidine (American Society of Health-System Pharmacists, 2001). Amiodarone, in particular, can increase serum digoxin concentrations 70 to 100 percent, with substantial variability. In the case report of a 76-year-old man provided by Castells et al. (2002) who reported reductions in visual acuity, color vision, increased difficulty with glare, and the presence of visual shining after the initiation of amiodarone, the patient’s cardiologist had reduced the dose of digoxin by the recommended 50 percent when amiodarone therapy was initiated.

The most common ocular symptom is halos or colored rings (usually blue-green) around lights, particularly at night. The incidence of colored rings or halos around lights ranges from 1.4 to 24 percent (Harris et al., 1983; Nielsen, Andraen, and Bjerregaard, 1983; Greene et al., 1983; Raeder, Podrid, and Lowan, 1985). The most common ocular finding is a type of keratopathy that consists of corneal epithelial microdeposits. A rare optic neuropathy, similar in ophthalmosopic appearance to nonarteritic ischemic optic neuropathy, has been described. The incidence of amiodarone optic neuropathy has been estimated at 1.8 percent versus a 0.3 percent risk of anterior ischemic optic neuropathy in an age- and location-matched group (Feiner et al., 1987). The severity of the vision loss from this optic neuropathy and the degree of its reversibility is variable (Harris et al., 1983; Mantyjarvi, Tuppurainen, and Ikaheimo, 1998; Ingram, 1983; Macaluso, Shults, and Fraunfelder, 1999; Sreith, Schoenfeld, and Marieb, 1999; and Feiner et al., 1987).

³⁷ Clinical Practice Guidelines: The Assessment, Management and Prevention of Falls in the Elderly in the South Western Sydney Area Health Service. An initiative of the Injury Advisory Committee. Michael Sugrue, FRCS (1), FRACS, Chair, Injury Advisory committee. www.swsahs.nsw.gov.au/livtrauma/practice.asp