Falls in older people are a common, dangerous and frequently incapacitating problem. They are often perceived as being untreatable—but this is an overly negative perspective. In any event, in the next few decades we will increasingly be confronted with elderly fallers as life expectancy continues to rise. This applies particularly to general practitioners, emergency department staff, geriatricians and neurologists. In this review, we will underscore the clinical significance of falls in the elderly and then outline a practical approach for their management. Core elements of this approach include:

• ascertaining whether or not the patient actually fell

• reliably classifying the nature of the falls

• identifying the causes and associated risk factors for falls

• tailoring an individualised treatment to the identified contributing factors, in order to reduce falls and fall-related injuries, or even to prevent them altogether.

WHY ARE FALLS IMPORTANT?

Falls in the elderly are a major health problem, first and foremost for the affected individuals whose quality of life is markedly reduced, and also for the public health system because of the immense costs associated with falls and the resultant injuries. The risk of falls increases with age: about one third of those over 65 years of age fall at least once a year, and about half of them even more often.¹ Apart from age, prominent risk factors include previous falls, female gender, concomitant neurological disease, living in a nursing home, fear of recurrent falling, and regular alcohol intake.^2,3


 About one third of those over 65 years of age fall at least once a year

Falling is serious, for several reasons:

• Falls may cause severe injury, and in up to 25% of elderly fallers this requires medical attention.⁴ Hip fractures are common and widely feared, and secondary complications due to immobility are frequent.

• Secondary immobility after a fall is common, and can be devastating in its own right as this promotes osteoporosis, which in turn increases the risk of fractures following future falls. A driving factor behind immobility is a fear of recurrent falls, which is regularly experienced by elderly fallers and may occur even after a single and seemingly innocent fall. For some patients, this fear of falling is appropriate because their balance is severely disturbed, but for others the degree of fear is disproportionate and leads to unnecessary immobility, loss of independence and even social isolation.

• Up to 50% of elderly fallers are unable to get up after a fall, not only as a result of injury, but more commonly because of physical frailty and proximal muscle weakness. Patients who lie on the ground for a long time may develop dehydration, pressure sores, rhabdomyolysis, hypothermia or pneumonia, all of which eventually may be fatal.

• Falling and fall-related injuries are a prominent reason for nursing home admission.

• Falls are often a marker for an underlying disease, progression of which may contribute directly to the increased mortality, for example in patients with cardiovascular or cerebrovascular disease.⁵

• Not surprisingly, quality of life among elderly fallers is markedly impaired.⁶

• Recurrent falls may reduce life expectancy, either directly (for example, subdural haematoma following head trauma) or indirectly due to complications of the fall.

A vicious circle of falling, balance problems, fear of falling, immobilisation and isolation mainly occurs in those who have recurrent falls, or who lie on the ground for a long time after their fall (fig 1).

THE CLINICAL APPROACH TO FALLING

Background

Many clinicians regard falls as an unavoidable accompaniment to normal “ageing”. However, up to 20% of very old individuals still have a completely normal gait and do not fall despite their age—indicating that balance and gait disorders are certainly not an inevitable consequence of ageing. Indeed, falls in the elderly should always initially be regarded as pathological and therefore require the identification of some underlying disease or risk factor. Falls tend to deter doctors because of their complex underlying pathophysiology, and clinicians are often frustrated in their approach to elderly fallers because their accounts, and even those of eyewitnesses, are often incomplete. Too often falls in the elderly are perceived as untreatable, and therefore deemed unsatisfactory to deal with. We hope this review will remove some of these false preconceptions so that clinicians come to regard elderly fallers as a gratifying challenge, rather than as a frustration.

Bedside history and examination

As ever, the approach to elderly fallers requires a thorough history (table 1), careful review of medical records, eyewitness reports and fall diaries, as well as a detailed physical examination.² Questions should not only focus on the falls per se, but also on their consequences; hip fractures are typically caused by lateral falls, bilaterally damaged patellas by drop attacks, and wrist fractures by a fall on the outstretched hand suggesting that consciousness was preserved while falling. The medical records should contain a lot of relevant and readily available information, including medical history, home circumstances and, importantly, use of sedative drugs or other predisposing medications.

Physical examination (table 2 and fig 2) should include a careful gait and balance assessment, preferably using “functional” tests which focus on the performance of everyday activities, a search for underlying risk factors, and any physical injuries.²

Evaluation of gait is mandatory because any walking problems increase the risk of falling: a shuffling gait increases the risk of stumbling over obstacles, and episodic gait disorders commonly lead to falls because patients are caught unprepared. For example, freezing of gait, where patients suddenly feel as if their feet have become glued to the floor⁷ is seen in Parkinson’s disease, as well as other parkinsonian disorders. Assessment of freezing is notoriously difficult because it is so often absent in the examination room. Helpful aids are the use of specific freezing questionnaires, interviewing the spouse, or demonstrating to the patient what freezing actually looks like. Useful tricks to provoke freezing include asking patients to initiate gait, to negotiate a narrow passage, and to walk while performing a “dual task” (for example, answering questions, or carrying an object). Perhaps the best test is to ask patients to turn around 360°, first in their preferred direction, and then in the opposite direction—the latter will provoke freezing much more often. Recognition of freezing requires insight into the three different clinical phenotypes: shuffling with small steps; “trembling in place” (rapid shuffling movements of the feet, but without the patient moving forward); or (more rarely) akinesia with complete inability to start or continue walking.

Standardised rating scales such as the Tinetti Mobility Index⁸ and the Berg Balance Scale⁹ are useful tools to describe most aspects of balance and gait, and we also value the opinion of an experienced physiotherapist who can review the patient more extensively in various circumstances, and also judge how well patients can use their walking aids.

Timed tests (for example, the 6-minute walking test, timed up and go test, single leg stance duration) have the advantage of providing quantitative information, and scores beyond established cut-off values may help to predict the risk of falls (table 2).¹⁰

Orthostatic hypotension is detected by measuring blood pressure, first in a recumbent position (preferably after a rest), and again after 1, 3 and 5 minutes of standing. Meanwhile, the patient should be observed for signs and symptoms of orthostatic hypotension such as “dizziness”, pallor, perspiration and stumbling. Note that clinically relevant orthostatic hypotension can be missed if the blood pressure is only measured once, and continuous blood pressure recording while patients are passively tilted upright may even be required.¹¹

Assessment of vision with and without correction is important because many falls are related to visual impairment. Paradoxically, poor vision is probably worse than no vision at all; poor vision provides false feedback and leads to incorrect movement planning, while no vision can at least be replaced by the intact remaining senses (for example, proprioceptive and vestibular) thanks to the physiological redundancy between these three afferent systems.

The need for cognitive testing is underscored by the accumulating evidence of a close relation between falls and cognitive decline (fig 3). For example, falls are related to being unable to walk and talk at the same time; inability to perform such seemingly simple tasks simultaneously has proven to be a good predictor for future falls.¹² Moreover, some elderly people loose their ability to deal with complex “multitask” circumstances and fail to give priority to the most important task (maintaining balance) at the expense of an increased risk of falling.¹³ Falls are common in patients with cognitive decline caused by Alzheimer’s disease (for example)¹⁴ and they are exceptionally common in disorders that combine motor impairment with cognitive decline, such as progressive supranuclear palsy. Cognitively impaired people are also more frequently exposed to dangerous situations due to their inability to estimate the risk of falling, and their loss of control of gait velocity.¹⁵ A global impression of cognitive function can be obtained during the history, and formal tests for frontal executive dysfunction are particularly important in patients with gait disorders.¹⁶ Additional bedside and sometimes more elaborate neuropsychological tests should be used to detect underlying conditions such as Alzheimer’s disease or vascular dementia.

Ancillary investigations

Quantitative gait and balance assessments (for example, static or dynamic posturography) are interesting scientific tools, but their use in daily practice is generally precluded by their high costs and insufficient individual diagnostic value.¹⁷ Brain imaging, preferably with MRI, should be considered in patients with unexplained falls, mainly to detect treatable disorders such as hydrocephalus, or disorders which might at least explain the falling even if they are not easily treatable such as periventricular white matter changes.^18,19 Other tests must be tailored to the clinical suspicions raised during the history and examination, and serve to identify underlying disorders. A more elaborate work-up may be required for patients whose falls were preceded by transient loss of consciousness:

• blood tests to detect electrolyte and glucose disturbances

• electrocardiography for suspected cardiac syncope, with 24-h monitoring if necessary

• electroencephalography when seizures are in the differential diagnosis

• carotid sinus massage to detect the carotid sinus syndrome

• tilt table testing to provoke syncope under controlled conditions is useful in the diagnostic work-up of patients with suspected orthostatic syncope (common in the elderly) or vasovagal syncope (rare in the elderly).

DID THE PATIENT REALLY FALL?

This question can sometimes be surprisingly difficult to answer, and it can help to have a definition for falls: “any sudden, unexpected event that caused the person to unintentionally land on any lower surface (object, floor or ground), regardless of any sustained injury”. Many older people do not mention their falls, simply because they accept falling as part of their ageing process. Others forget their falls, even if injury occurred, partly because of the association between memory impairment and falling.²⁰ And even if patients do recall their falls, they often find it difficult to recall exactly under what circumstances the fall occurred.² Asking patients and/or their carers to keep a dedicated falling diary is useful to record the number and circumstances of the falls,²¹ and a “falls hotline”, where patients can report and discuss their falls immediately after they have occurred, is also helpful.

Whether a fall was preceded by loss of consciousness can be difficult to ascertain because syncopal falls are often associated with amnesia for the fall, even in cognitively intact elderly people. A useful trick is to ask the patients whether they recall hitting the ground after their fall. If not, consider transient loss of consciousness, even when patients deny it. Eyewitness accounts should be helpful, but are commonly unavailable or incomplete.²²

WHAT SORT OF FALL(S) DID THE PATIENT HAVE?

Once it has become clear that a patient really has fallen, the next step is to classify the nature of the fall(s), which provides the basis for tailored treatment (fig 4).

A single fall or recurrent falls?

This is important, because single falls with an obvious extrinsic cause (like ice on the pavement) merely require treatment of any associated injuries, without further analysis into the cause of the fall. But if no obvious extrinsic cause can be found, or if a patient has had recurrent falls, further investigations are justified.

Is there a pattern to the falls?

For patients with recurrent falls, the next step is to identify any stereotypical pattern. For example, patients may say their falls occur exclusively immediately after rising from sitting or lying, and so they could have orthostatic hypotension, or severe balance impairment leading to insecure transfers. Identification of fall patterns also has therapeutic consequences. For example, adapting the house is a useful approach for patients who only fall due to trips over doorsteps, or while climbing stairs. When patients present with different types of falls, each fall type should be scrutinised separately.

Were extrinsic or intrinsic risk factors (or both) involved?

The next step is to decide whether the falls were predominantly related to “intrinsic” (patient-related) factors (yellow boxes in fig 3) or “extrinsic” (in the environment) factors (orange box in fig 3). And this process should also focus on identifying specific protective factors (which will be discussed in the following paragraph), because this determines the net risk of falling (green box in fig 3). Note that falling in the elderly is typically a multifactorial problem, where multiple risk factors jointly contribute to falls in each individual patient. And that falling often results from interactions between intrinsic and extrinsic factors. For example, a doorstep might only create problems when step height is diminished, as in patients with Parkinson’s disease or a dropped foot due to ankle extensor weakness. Therefore, physicians should not stop when a single risk factor has been identified, but instead pursue a systematic search for multiple intrinsic and extrinsic risk factors, as well as any protective factors.


 Falling in the elderly is typically a multifactorial problem, where multiple risk factors jointly contribute to falls in each individual patient

Extrinsic (environmental) factors

Extrinsic risk factors include freshly polished floors, wet bathroom tiles, stairs, loose carpets, uneven pavements, poor lighting, stepping onto escalators, and dogs or cats in the household.³ Inappropriate footwear (high heels, slippery soles or loosely fitting shoes) is another common extrinsic factor. The risk of falling indoors is also associated with walking barefoot, or in socks. Modern buses and trains with their fast acceleration and automatic doors can cause considerable difficulty for elderly people who may fall before they can find a seat.

Intrinsic (patient-related) factors

Many elderly people cannot identify clear extrinsic determinants for their fall, and have repeated falls in seemingly harmless situations. They merit a thorough work-up of intrinsic risk factors as there is a high risk of recurrent falls.² Intrinsic risk factors often include one or more underlying disorders, in combination with drugs, alcohol or both (table 3; fig 3).

Use of medication is a prominent risk factor in the elderly. The underlying pathophysiological mechanisms may include a combination of sedation, cognitive impairment, carotid sinus syndrome, orthostatic hypotension, urinary incontinence, behavioural abnormalities, extrapyramidal adverse effects, ataxia, and muscle weakness. Particularly notorious are benzodiazepines and antidepressants,²³ recent initiation of new medication and polypharmacy. Neuroleptics, antihypertensive medication, and anti-arrhythmics also increase the risk of falls.²⁴ In Parkinson’s disease, dopaminergic medication may paradoxically increase the falling frequency by causing violent dyskinesias, sudden freezing of gait, orthostatic hypotension or confusion.²⁵

Many chronic diseases are associated with falls (table 3), both acute disorders (for example, delirium, urinary tract infection with urge incontinence) and a wide range of chronic conditions. Note that physical impairments such as urine incontinence or visual impairment are more important than the diseases themselves in predicting recurrent falls. One important example is diabetes mellitus, which may contribute to the risk of falling by various mechanisms, including hypoglycaemia, diabetic retinopathy, polyneuropathy, foot ulcers and stroke. Urge incontinence is also associated with falls, partly because the underlying disease may cause incontinence and falls (for example, stroke), and partly because night-time visits to the toilet in darkness provide ideal falling circumstances.^26,27 Osteoporosis should be suspected in patients who have low-impact fractures, in nursing home residents, in frail elderly people and those taking steroids.

Were the falls associated with a transient loss of consciousness?

For patients with intrinsic falls, the next step is to clarify whether they were preceded by transient loss of consciousness (distinguished from loss of consciousness occurring after the fall—for example, due to head injury). If so, their diagnostic and therapeutic work-up is completely different from patients who fell with preserved consciousness; epilepsy, syncope and psychiatric disorders all need to be considered (table 4).

A major category is syncope, where—by definition—loss of consciousness is caused by failure of the cerebral circulation, itself almost always due to a failure of the systemic blood circulation. The falls are mostly backwards, either flaccidly or stiffly. Syncope is typically preceded by presyncopal features, such as blurred vision and loss of colour vision (“greying out”), loss of control over eye and other movements, constriction of the field of vision, and hearing loss. The loss of consciousness usually lasts less than 20 seconds, and involuntary multifocal jerky movements may be seen, but in contrast with epileptic seizures, these are non-rhythmic and asynchronous involving various body parts.¹¹ Urinary incontinence is common (so this does not differentiate syncope from epilepsy) but a lateral tongue bite is rare (this does suggest epilepsy). Patients recover promptly and spontaneously, with rapid improvement of behaviour and cognition. In the elderly however, syncope can present in a less typical way without all the usual features, and retrograde amnesia is common. Orthostatic hypotension and the carotid sinus syndrome are the most common causes of syncope in the elderly,^28,29 and it is most important to identify cardiac syncope because mortality is doubled due to the underlying heart disease (we refer to a recent review for further details of syncope¹¹).

Epilepsy is a less common cause of falls in the elderly. Eyewitness reports are crucial for the diagnosis. The characteristic tonic-clonic with massive, synchronous jerking movements of face and limbs usually occur after the patient has fallen to the floor. The duration of the actual seizures is generally brief, often only a few minutes. However, unlike syncope, postictal confusion and anterograde amnesia are distinctive features, typically for at least several minutes, but often much longer. Seizures in old age call for a search for an underlying cause, such as a brain tumour or other focal lesion. Note that epilepsy is sometimes the result of a fall, rather than its cause—for example, in the case of brain concussion or intracranial haematoma. Electroencephalography may confirm the clinical suspicion of epilepsy and help classify the seizure type, but it is often normal in the inter-ictal phase. More importantly, EEG can assist in predicting the risk of recurrent seizures.

Was the fall preceded by a “funny turn”?

When preceding loss of consciousness is ruled out, the next step is to ascertain whether the fall was preceded by a “funny turn” of some kind. Many elderly people report “dizziness” as the cause of their falls. Try to clarify precisely what they mean: was it vertigo (spinning sensations, usually accompanied by nausea) or light-headedness (perhaps with presyncopal features)? Ask patients to specify whether their dizziness was “in the head” (suggesting syncope or vertigo), “in the eyes” (suggesting poor vision) or “in the legs” (suggesting venous insufficiency, or sensory ataxia due to polyneuropathy). Note that falls preceded by funny turns do often turn out to be caused by syncope, even when loss of consciousness is denied.

Is there any balance or gait disturbance?

If the fall was the result of an intrinsic factor, but not preceded by loss of consciousness or some funny turn, the next question is whether the patient’s balance or gait is sufficiently disturbed to explain it. If these are normal or only minimally impaired, falls might be caused by drop attacks or, in younger people, by cataplexy or hyperekplexia.¹¹ Note that patients may have episodic gait disturbances that are easily missed in the office, but which are notorious causes of falls because their episodic nature makes it difficult for patients to adapt their walking behaviour—for example, freezing of gait in parkinsonian disorders, and neurogenic claudication due to lumbar stenosis.⁷

“Base-of-support” or “centre-of-mass” falls?

Falls caused by balance or gait disturbances can be subdivided into two categories:

• “base-of-support falls” caused by displacement of the feet (for example, slips, or tripping over obstacles)

• “centre-of-mass falls” related to trunk instability, either during self-induced perturbations (for example, bending, reaching or turning) or caused by externally applied perturbations (for example, a push or collision).³⁰

Making this distinction helps to identify the underlying problem: base-of-support falls occur in patients with lower leg weakness (dropped foot) or distal sensory loss caused by polyneuropathy. Centre-of-mass falls are common in parkinsonian patients. This distinction also has implications for treatment, because frequent base-of-support falls could justify a home visit by the occupational therapist to remove obstacles from the floor. Conversely, centre-of-mass falls call for optimised treatment of the underlying balance deficit in trunk control.

What if patients think they fell spontaneously?

What to do when the physician is unable to pinpoint even a single risk factor for apparently spontaneous falls is a common problem:

• One diagnosis to consider is drop attacks, typically in middle-aged women who fall spontaneously during walking, and only rarely while just standing, without loosing consciousness.¹¹ Men are rarely affected. The typical story is of sudden, unexpected falls without preceding loss of consciousness, and without prodromal or postictal symptoms. There is no relation to change in posture, head movement or any other specific precipitating event. Patients typically fall straight down or forward onto their knees. Unlike syncope and epilepsy, there are no associated involuntary movements. If no injury occurs, the patients can get up immediately and resume their activities. Most of these drop attacks remain cryptogenic.¹¹

• A second possibility is that the patient did in fact fall due to loss of consciouseness, but failed to recognise this (see above).

• A third option is freezing of gait, which we described above. Not all patients can properly identify their freezing episodes, but instead report “spontaneous” falls.

Asking about the direction of falls can provide a diagnostic clue: forward falling suggests freezing of gait in Parkinson’s disease; (sudden) backward falling suggests progressive supranuclear palsy; and vertical (downward) falling suggests syncope or drop attacks. Lateral falls have little diagnostic value, but have great clinical relevance because these cause hip factures. Inability to stand up after a fall may suggest a preceding epileptic seizure or stroke.

HOW TO PREVENT RECURRENT FALLS

Prevention of falls and subsequent injuries requires treatment of any underlying disorder and elimination of the associated risk factors.

Primary prevention focuses on elderly people who have not yet fallen, and aims to eliminate risk factors that are common in the elderly such as lack of exercise, or unnecessary use of psychoactive drugs. Tackling risk factors that are only weakly associated with falls, such as inappropriate footwear, may still be rewarding if they are sufficiently prevalent in the general population. Preventing osteoporosis reduces the chance of fracture should the patient fall. Various exercise programmes—walking, Tai Chi and dancing—can clearly improve strength, endurance and balance, and several controlled trials have shown a significant reduction in falling.³¹

Secondary prevention focuses on elderly people who have fallen at least once and aims to avoid recurrent falls. Here, the emphasis is more on treatment of specific underlying disorders and eliminating intrinsic or extrinsic risk factors that are strongly associated with falls.

Tertiary prevention concerns measures that benefit elderly people who fall very often, have sustained recurrent injuries, and who have risk factors for falls that are hardly amenable to secondary prevention. This includes frail elderly people in nursing homes, demented patients, and patients with severe motor handicaps—for example, end-stage parkinsonism. Here the aim is to limit the impact of falls using hip pads, installing personal alarm systems, or home surveillance of solitary elderly people.³²

Falls prevention should involve a multidisciplinary falls team,³³ including a clinician with appropriate skills and experience or a specialist (for example, geriatrician, neurologist), physiotherapist, occupational therapist and specialist nurse. The proposed interventions should be practical and easy to implement, and this calls for optimal cooperation with the patient and their family. Patients should receive both verbal and written information on the preventive measures. Table 5 lists possible strategies for secondary prevention.^32,34,35

Several measures deserve specific attention:

• Drug combinations of any type should be avoided in elderly fallers, and the pharmacist can play an active role here. A recent randomised controlled trial showed that when a pharmacist reduced the number of drugs in elderly care home residents, the number of falls was reduced by 40%.³⁶ Psychoactive drugs (mainly benzodiazepines and antidepressants) should be stopped or minimised. Drugs in high doses or with long half-times should particularly be avoided.

• Promoting the use of walking aids also deserves specific attention. Many elderly people do not use them, either because they are not recommended to do so or because they are ashamed or embarrassed. For many older people, the use of walking aids can be unsafe but if those at risk are instructed properly, they are often pleased with their regained confidence, mobility and independence. Physiotherapists or occupational therapists can assist in selecting the appropriate walking aids, and train patients how to use them properly.³⁷ Physiotherapists can also teach people to make safer transfers and increase their physical fitness.

• Home visits by occupational therapists can reduce domestic hazards. In addition, medical and behavioural risk factors for falls can be screened at the same time, and replacing unsafe footwear also helps to reduce the number of falls. However, poor compliance limits the success of home visits. Many elderly are reluctant to change their house or their behaviour, or doubt that doing so will reduce their fall frequency. Therefore, their agreement to any preventive measure is essential.³⁸ Another aspect to consider is that home visits are generally costly and time consuming.³⁹

• Prevention of falls associated with loss of consciousness mainly involves treatment of the underlying disorder.¹¹

  • Orthostatic hypotension patients should be advised to avoid precipitating circumstances, such as rapid changes in posture or prolonged recumbence. In addition, they can be trained to use anti-orthostatic manoeuvres such as standing with crossed legs or regular “tip-toeing”, or elastic compression stockings. A simple pharmacological intervention is sodium chloride tablets. If this fails, fludrocortisone and sympathomimetics can be used, but only when there are severe disturbances of blood pressure regulation, because of their adverse effects in the elderly.

  • Elderly fallers with urge incontinence should be scrutinised for vascular dementia, spinal cord compression, or normal pressure hydrocephalus.

  • Proper lighting should be installed for those who frequently fall during night-time visits to the toilet. A commode next to the bed or a condom catheter for men may be needed.

  • For benign paroxysmal positional vertigo, an Epley manoeuvre (to eliminate the debris from the semicircular canals) may dramatically improve the debilitating complaints, but is often difficult to perform in elderly people. When this manoeuvre fails, habituation exercises can be tried. There are no drugs with proven anti-vertigo efficacy.

  • Prevention of osteoporosis is recommended for elderly fallers. Promoting physical activity increases bone mineral density and reduces the risk of hip fractures. Oral provitamin 1α-hydroxyvitamin D3, calcium supplementation and alendronate/raloxifene arrest progression of osteoporosis and reduce hip fractures.⁴⁰

Tertiary preventive measures focus on the fear of falling and the resulting immobilisation or social isolation. An electronic warning system around the neck or wrist can limit complications in patients who are unable to stand up after a fall. Fear of falling can of course be reasonable in patients with severe balance impairment, and the resultant restriction of mobility can actually serve as an adequate tertiary preventive measure. However, for many, the fear is disproportional to the actual degree of balance impairment and risk of falls. Reduction of fear and regaining confidence is important for these people, because it helps restore mobility and promotes independence. Group treatment using a behavioural-cognitive approach to change attitudes, as well as training with a physiotherapist, might also help. In addition, physiotherapists play a central role in restoring balance confidence and reducing the fear of falls.⁴¹ However, this positive effect wanes as the patient becomes more frail.⁴² An entirely different approach is required for cognitively impaired patients who can be too confident and inappropriately overrate their own balance ability, resulting in risky behaviour and falls. For them, restriction of activities might be the best solution to prevent recurrent falls.