Essential Tremor: Diagnosis and Treatment

Jack J. Chen, PharmD, David M. Swope, MD


Pharmacotherapy. 2003;23(9) 

In This Article


In one community-based study, only 11% of patients with essential tremor had sought medical advice for tremor.[41] Of those who do seek medical advice, approximately 15% do so only after significant functional disability has occurred.

The goal of treatment is to minimize functional disability, reduce social handicap, and improve quality of life. The selection of treatment options is based largely on the patient's needs and history (tremor severity, coexistent disease, current drug therapy, response to previous therapy). Treatment may include physical therapy, behavioral and psychologic interventions, lifestyle changes, pharmacotherapy, and surgery. Drug therapy does not cure, prevent, or slow the rate of disease progression and is considered symptomatic treatment. Treatment is not required if the patient does not have significant functional or psychosocial disability. Figure 2 is a treatment algorithm, and Table 7 provides dosages of common antitremor drugs.

Treatment algorithm for essential tremor. Vim = ventralis intermedius nucleus of the thalamus; DBS = deep brain stimulation.

Most patients with mild essential tremor are able to minimize functional disability, social embarrassment, and personal injury by learning adaptive techniques. Examples include learning to write with the least disabled hand, placing a napkin between cup and saucer to avoid rattling, avoiding difficult foods (e.g., soup, spaghetti), using blunt-tip safety scissors, wearing clip-on neckties, having autodial on a telephone or asking the operator to place calls, learning deep breathing and other relaxation techniques, avoiding awkward or uncomfortable situations, and explaining their condition to people. The number of adaptive techniques is numerous, and people can be very creative.

Physical and occupational therapists may offer suggestions regarding wrist weights, plate guards, and other adaptive devices. These devices can provide considerable benefit in activities of daily living. Additional sources of information for clinicians and patients can be found by contacting the International Tremor Foundation (7046 West 105th Street, Overland Park, KS 66212-1803;

For patients with mild tremor, minimizing exposure to emotional stress, tremorogenic foods and drugs, and reassurance are all that is required. Intermittent administration of a β-blocker or small amounts of alcohol may be effective in special social situations.

When tremor significantly interferes with daily activities, long-term pharmacotherapy is indicated, with drug selection dictated by comorbid conditions and anticipated safety and efficacy. The primary goal is to minimize drug side effects while providing maximum improve-ment in function. Patients should be informed that although significant benefit may be derived, complete tremor eradication is not a realistic expectation.

Drugs reduce tremor amplitude and associated disability. Tremor frequency is not significantly affected and is generally not correlated with disability. Assessment techniques for drug effectiveness include tasks such as handwriting, pouring, and drinking. It is important to recognize that improvements in accelerometric measurements (tremor amplitude) may not translate into proportional improvements in function. In addition, patients may improve in certain tasks such as drinking from a cup or using food utensils, but tasks requiring fine hand manipulations may not improve.

β-Blockers. Therapy with propranolol for management of essential tremor dates back over 40 years.[4] Since then, the drug's efficacy was confirmed in several double-blind, placebo-controlled, crossover trials, supporting its status as one of the most widely administered agents for management of essential tremor symptoms.[42] Propranolol is the only agent that has approved labeling for this indication.

Lipid solubility does not appear to be critical for tremorlytic activity, as water-soluble agents (e.g., nadolol) are also effective.[43] β1-Receptor selective drugs (e.g., atenolol, metoprolol) are also effective, but β2-receptor blockade appears to be required and the efficacy of selective β1-receptor blockers may be due to spillover β2-receptor antagonism.[44,45] Absence of intrinsic sympathomimetic activity appears to be important because pindolol is ineffective.[46] Of β-blockers, propranolol appears to be the most effective, and lack of response to it generally predicts lack of response to other drugs in this class.

The exact tremorlytic mechanism of propranolol remains unknown but may involve blockade of peripheral β2-receptors within muscle fibers or muscle spindles.[47] Approximately 50-60% of patients experience some improvement in functional disability due to a reduction in tremor amplitude, but total tremor suppression rarely is achieved. The greatest improvement is in hand tremor and the least in head and voice tremors.[48] The duration of effect after each dose is approximately 4 hours, so two daily doses are required. Sustained-release propranolol may enhance adherence and therefore therapeutic outcomes. In 18 patients with essential tremor, sustained-release propranolol was compared with the regular-release formulation and was preferred by 67% for tremor suppression and by 87% for ease of administration.[49]

The short- and long-term tolerability of propranolol were evaluated in a nonrandomized, comparative study of 25 patients who received long-acting propranolol 80-160 mg/day and 25 who received primidone 50-250 mg at bedtime.[50] Patients were followed for 12 months. Tremor measurements included a clinical severity score based on observable tremor amplitude, ability to perform writing, drawing (spirals), and pouring tasks, and linear accelerometry. Of patients receiving propranolol, the mean age was 68.9 years and mean disease duration was 20.4 years. Approximately 1 in 10 of them experienced short-term side effects (bradycardia, syncope) and 1 in 5 experienced long-term side effects (bradycardia, fatigue, erectile dysfunction) that required discontinuation of therapy.

When taking propranolol, monitoring of heart rate and blood pressure is recommended, especially during dose titration. Side effects such as diminished exercise tolerance, fatigue, bronchospasm, depression, insomnia, and erectile dysfunction may not be tolerated by some patients. Those with underlying cardiovascular disease should not have long-term propranolol therapy discontinued abruptly due to possible rebound cardiovascular events. Relative contraindications include severe heart failure, cardiac conduction blocks, and bronchospastic conditions. Propranolol also can inhibit sympathetically mediated hypoglycemic response in patients with diabetes, and benefits of therapy should be weighed against risks.

Propranolol may be administered on an intermittent (as needed) or scheduled basis. When administered intermittently, 20-mg tablets are preferred, with instructions to take one-half to two tablets about 30 minutes to 1 hour before social activities or anxiety-provoking events that may increase tremor. If administered as long-term suppressive therapy, it should be started at 20 mg twice/day. Optimal tremor reduction usually is achieved at 120-240 mg/day. Higher dosages may be given, but dosages greater than 320 mg/day do not appear to provide additional benefits. For elderly and frail patients, treatment should be begun at 10 mg twice/day and gradually titrated to an average of 80-120 mg/day. If propranolol is effective and tolerated, the once-daily long-acting preparation may be substituted.

Generally, propranolol retains its antitremor effect during long-term therapy. However, after 1 year, approximately 10-15% of responders develop tolerance to the tremorlytic effect and may require dosage increases.[50] β-Blockers should be considered as first-line therapy unless the patient has contraindications or history of β-blocker intolerance. Concomitant conditions that may benefit (e.g., hypertension, stable angina, mild-to-moderate chronic heart failure, migraine headache) would strengthen the selection of a β-blocker.

Primidone. Over 20 years ago, primidone, a structural analog of phenobarbital, was serendipitously discovered to have antitremor activity when it was given to a patient with epilepsy and essential tremor.[5] Since then, its efficacy for essential tremor has been confirmed by several double-blind, placebo-controlled, crossover trials.[42] The principal metabolites of primidone, phenobarbital and phenylethylmalonamide (PEMA), do not possess significant tremorlytic activity. Therefore, primidone or an unknown metabolite appears to be the active tremorlytic agent. The exact mechanism of action remains unknown, and serum concentrations of primidone, PEMA, and phenobarbital are not correlated with antitremor efficacy.[51,52]

Primidone is similarly or slightly more efficacious than propranolol, with near complete tremor suppression achieved in a greater proportion of patients.[50] However, as with propranolol, primidone is most beneficial for essential hand tremor, and efficacy against head and voice tremor is variable. The duration of effect after a single dose is approximately 24 hours. As with propranolol, after 1 year of therapy approximately 10-15% of responders may develop tolerance to the tremorlytic effect of primidone.

In the short term, primidone is less well tolerated than β-blockers. In the nonrandomized, comparative study of 50 patients who received either propranolol or primidone, of 25 patients receiving primidone, the mean age and tremor duration were 66.6 years and 20.4 years, respectively.[50] Nearly one-third of patients experienced short-term side effects (nausea, ataxia, dizziness, sedation, confusion, malaise) that occurred the morning after the first nighttime dose and persisted for up to 4 days; however, only 12% discontinued therapy. All patients had been educated as to the potential for reactions and that they would abate rapidly. During long-term therapy, only 8% of patients complained of side effects (sedation) and none discontinued the drug due to side effects.

Primidone is contraindicated in pregnancy, breastfeeding, and porphyria, and caution should be exercised in debilitated patients and those with impaired hepatic or renal function. Serious complications are rare and include red cell hypoplasia or aplasia, agranulocytosis, and megaloblastic anemia. A complete blood count should be performed at baseline and every 6-12 months to screen for blood dyscrasias. Clinicians also must keep in mind the potential for drug interactions secondary to the phenobarbital metabolite.

Primidone is available as scored 50- and 250-mg tablets and as an oral suspension of 250 mg/5 ml. It typically is prescribed on a constant-use basis. Because it has a long duration of antitremor activity, dosing is once/day. Strategies to minimize short-term reactions include starting therapy at a subtherapeutic bedtime dose (12.5 mg). The suspension formulation is particularly helpful for administering small doses. The dosage should be titrated slowly upward for desired tremor control. Most patients achieve optimal benefit with dosages of 250 mg/day or less. If higher daily doses are required, administration in several daily doses is recommended. Strategies to promote adherence to primidone include educating patients on potential short-term side effects and reassuring them that these reactions will disappear after the first few doses, and attempting to achieve optimal symptomatic control with a single daily dose.

The role of primidone is as a second-line agent for monotherapy or first-line agent in patients who do not tolerate propranolol. Primidone also may be used as an add-on drug if symptomatic relief is insufficient despite maximally tolerated doses of existing monotherapy.

Gabapentin. The efficacy and safety of gabapentin for the management of essential tremor were evaluated in several randomized controlled trials.[53,54,55] Although the drug's antitremor effect remains unknown, enhancement of central GABAergic tone may play a role.

The efficacy of gabapentin 1800 mg/day in divided doses for essential tremor was evaluated in a randomized, double-blind, placebo-controlled, crossover trial.[53] Clinical tremor outcomes were measured by the Fahn-Tolosa-Marin Tremor Rating Scale and quality of life by the Sickness Impact Profile (SIP). Treatment duration was 2 weeks with a 5-day washout between active drug and placebo. All 20 patients had disabling essential hand tremor (mean age 66.5 yrs; mean tremor duration 33.4 yrs). Renal function parameters were not reported. At the conclusion of the study, subscales of the tremor-rating scale (tremor severity, motor task performance, activities of daily living) and SIP scores were no different from baseline in both groups. However, this study was not designed to assess the efficacy of gabapentin monotherapy, as 70% of patients were taking one or more tremorlytic agents (e.g., propranolol, primidone, clonazepam, metha-zolamide) during the study.

A randomized, double-blind, placebo-controlled, crossover study compared gabapentin and propranolol.[54] Sixteen patients received monotherapy with divided doses of gabapentin 1200 mg/day, propranolol 120 mg/day, and placebo for 14 days with a 1-week washout period between treatments. All patients (mean age 67.9 yrs; mean tremor duration 12.2 yrs) had postural and kinetic hand tremor. Although none of them was taking other antitremor agents, 10 (62.5%) had received propranolol or diazepam in the past. Renal function parameters were not reported. Tremor was evaluated by the Fahn-Tolosa-Marin Tremor Rating Scale and linear accelerometry. Compared with placebo, gaba-pentin and propranolol were associated with statistically significant improvements in all subscales of the tremor-rating scale and patients' subjective assessment. Although not statistically significant, tremor-rating scores were slightly better for propranolol than for gabapentin. When asked which treatment was more effective, however, 50% of patients indicated gabapentin, 38% propranolol, and 12% both. As would be expected, a number of patients (~38%) did not respond to either drug, and approximately 19% responded positively to placebo.

Electromyographic (EMG) recordings from the extensor and flexor forearm muscles also were obtained and suggest that pharmacologic response may correlate with various EMG patterns. In patients with simultaneous contractions in antagonist muscles, propranolol was more effective and in those with alternating contractions, gabapentin was more effective. Although EMG measurements may not be practical in routine office practice, confirmation of these results may yield insight into the pharmacodynamic mechanism of antitremor agents.

A randomized, double-blind, placebo-controlled, crossover, dose-escalation study assessed the efficacy of gabapentin 1800 mg/day and 3600 mg/day for control of essential hand tremor.[55] Among 20 patients completing the study, mean age was 69.9 years and mean tremor duration 29.1 years. This study was not designed to assess the efficacy of gabapentin monotherapy because most patients also were taking one or more tremorlytic agents (primidone, propranolol, benzodiazepine). After a 1-week titration phase, patients were maintained with gabapentin 1800 mg/day or placebo in divided doses for 2 weeks and evaluated. Doses were titrated over another week to 3600 mg/day. After 2 weeks at 3600 mg/day or placebo, another evaluation was performed. After a 1-week washout period, patients receiving gabapentin were switched to placebo and vice versa, and the 6-week evaluation process was repeated. Tremor severity was measured by the Unified Tremor Rating Scale and triaxial accelerometry. No statistically significant difference in accelerometry measurements was seen between placebo and the two gabapentin dosages. However, not surprisingly, this did not correlate with functional disability scores, and gabapentin 1800 mg/day was associated with statistically significant improvements in scores for activities of daily living and pouring; however, improvements in spiral drawing were not significant. Similar, but not superior, results were associated with high-dose gabapentin.

Interpretation of the results of these trials suggests that short-term treatment with gabapentin 1200-1800 mg/day is well tolerated in elderly patients with essential tremor and that the drug is effective as monotherapy for essential hand tremor. As add-on therapy to standard antitremor agents, the benefits are less robust and variable. Gabapentin is associated with few drug interactions, and long-term therapy appears well tolerated in both young and elderly patients, although ataxia, irritability, sedation, and weight gain may occur. Additional clinical studies are required to determine the agent's true benefit; however, due to a favorable safety profile and ease of use, gabapentin may be considered an alternative second-line agent for management of essential tremor and also as an add-on agent if symptom relief is insufficient despite maximally tolerated dosages of concurrent antitremor therapy.

Alcohol. The tremorlytic activity of alcohol in essential tremor has long been recognized, with up to two thirds of patients reporting temporary relief.[15] Studies of positron emission tomography reveal that alcohol reduces the cerebellar hyperactivity in essential tremor.[56] Although not confirmed, this may be mediated by GABAergic mechanisms. A glass of wine or light cocktail is often enough to attenuate the tremor for up to an hour; however, a rebound exacerbation of tremor often occurs. Many patients have learned to self-medicate in a controlled manner, and although it was suggested that they have an increased risk of developing alcoholism, their pattern of alcohol use is the same as that of the general population.[57] Although routine ingestion of alcohol for symptom control cannot be widely proscribed to patients with essential tremor, many do benefit from the sparing and responsible ingestion of alcohol before selected tasks or social events.

Benzodiazepines. Benzodiazepines, such as alprazolam, clonazepam, diazepam, and lorazepam, may be considered adjunctive therapy for patients whose essential tremor is not well controlled with standard agents. The drugs' tremorlytic mechanism of action is unknown but may be related to GABAergic augmentation resulting from interaction at the benzodiazepine receptor-GABA receptor-chloride channel complex.[47] Alternatively, the agents relieve or prevent essential tremor exacerbated by emotional stress or anxiety. Diazepam commonly is prescribed, but its efficacy is largely anecdotal as clinical studies have not been conducted. In a double-blind, placebo-controlled study, 22 elderly patients each received 4 weeks of alprazolam, acetazolamide, primidone, and placebo, in a crossover manner separated by 2-week washout intervals.[58] Alprazolam and primidone were similarly effective and both were more effective than placebo. Acetazolamide was no more effective than placebo. The mean effective alprazolam dosage was 0.75 mg/day, and the drug was well tolerated. Clonazepam is particularly effective for orthostatic tremor, a variant of essential tremor, and should be started at 0.25 mg/day and slowly increased over several weeks to 1.0-6.0 mg/day in divided doses.

Because of their anxiolytic properties, benzodiazepines may be more beneficial in frequently anxious patients, and small doses (lorazepam 0.25-0.5 mg) may be prescribed judiciously (as needed) and administered 30 minutes to 1 hour before an important event. However, due to central side effects (sedation, confusion, memory loss) and an increased risk of falls, benzodiazepines must be taken with caution by the elderly. Lorazepam may be preferred due to its mild sedative profile.

Botulinum Toxin. Botulinum toxin is the most potent biologic toxin known and is a powerful therapeutic tool in certain hyperkinetic movement disorders. Seven serologically distinct neuroparalytic toxins (types A-G) have been derived from Clostridium botulinum, and although they have similar molecular structures, each differs in pharmacologic characteristics. In the United States, botulinum toxin type A (BTX-A) is available as Botox (Allergan, Irvine, CA) and subtype B (BTX-B) is available as Myobloc (Elan Pharmaceuticals, South San Francisco, CA). It is important to note that, in addition to Botox, another BTX-A preparation is available outside the United States (Dysport; Ipsen, Berkshire, United Kingdom). These two preparations possess different potencies (Botox 20 U/ng, Dysport 40 U/ng). However, although the relative potencies are different, the clinical activity of 1 U Botox is equivalent to approxi-mately 3-4 U Dysport.

Three steps are involved in the paralytic mechanism of BTX-A.[59] The toxin is composed of two chains, heavy and light, linked by a disulfide bond. The heavy chain binds to a specific membrane acceptor on the presynaptic cholinergic terminal of the neuromuscular junction. Endocytosis of the light chain occurs. Finally, the light chain targets and cleaves the 25-kDa synaptosome-associated protein, a protein involved in the fusion of acetylcholine vesicles at the presynaptic membrane. The release of presynaptic acetylcholine is inhibited without impeding the synthesis and storage of acetylcholine. The mechanism is similar for BTX-B except that it prevents release of acetylcholine by cleaving a different cytosolic protein, synaptobrevin-2 or vesicle-associated membrane protein.

The size of the denervation field is determined by dose and volume (10 U BTX-A diffuses up to 4.5 cm). Once paralyzed, nerve terminals begin to form temporary neuronal sprouts and eventually motor neuron function is completely restored with reinnervation of the parent terminal and degeneration of auxiliary sprouts. Typically, a single treatment involves multiple point injections, and retreatment is required approximately every 3-4 months. The most common adverse effect to be expected is focal weakness due to unwanted diffusion of toxin into adjacent muscles (injections in the hands are associated with finger weakness, in the larynx with dysphagia). When administered parenterally, the lethal dose for 50% of organisms for Botox is approximately 40 U/kg or 3000 U for a 75-kg individual. Therapeutic dosages represent 1-10% of this lethal threshold.

Data from open-label and randomized, controlled studies support BTX-A in selected patients with hand, head, vocal, or palatal tremor.[60,61,62,63] In a randomized, double-blind, placebo-controlled trial, 25 patients with moderate-to-severe essential hand tremor were injected with BTX-A 50 or 100 U into wrist flexors and extensors and followed for 16 weeks.[60] Based on accelerometric data and ratings of tremor severity scales, 75% of BTX-A-treated patients experienced mild-to-moderate improvement 4 weeks after treatment compared with 27% of placebo-treated patients (p<0.05). No significant functional improvement was seen, however, because BTX-A injections in the hand also result in hand weakness. Greater improvement was noted for postural tremor compared with kinetic tremor. These results were confirmed in a randomized, double-blind, controlled trial involving 133 patients with essential tremor who were injected with BTX-A 50 or 100 U into wrist flexors and extensors and followed for 16 weeks.[64] Improvements in motor task performance and functional disability were inconsistent and appeared to be offset by the adverse effect of dose-dependent hand weakness. After long-term experience, many clinicians now give reduced doses of 15 U or less in forearm extensors to reduce tremor and to minimize finger extensor weakness. Because voice and head tremors are often resistant to oral agents, BTX-A often is considered the drug of choice for them, and in experienced hands, the risk of dysphagia after cervical or laryngeal muscle injection is reduced significantly.

Contraindications to botulinum include myasthenia gravis, post-polio syndrome, Eaton-Lambert syndrome, motor neuron disease, aminoglycoside antibiotics, and pregnancy. Serious systemic side effects (generalized botulism-like syndrome) from local BTX are rare.[65] A more common systemic effect is formation of neutralizing antibodies after several years of repeated injections. Because this results in therapeutic failure, steps to minimize antibody formation are important and include administering the minimum effective dose and extending the retreatment interval as long as possible, typically no earlier than every 3 months.

Miscellaneous Agents. Many other agents have been studied for the treatment of essential tremor, and preliminary results from open-label and controlled studies suggest that acetazolamide, clonidine, clozapine, methazolamide, mirtazapine, nicardipine, nimodipine, quetiapine, theophylline, and topiramate may be effective.[58,66,67,68,69,70,71,72,73,74] Clinical data for topiramate are encouraging. Results of double-blind, placebo-controlled, crossover trials suggest that topiramate up to 400 mg/day monotherapy or in combination with antitremor agents significantly improves tremor scores and tremor-related functional disability scores.[74,75] However, paresthesias, cognitive impairment, and anorexia may limit therapy.

Surgical interventions should be considered for selected patients with disabling tremor that is not adequately controlled with pharmacotherapy. Improvements in neuroimaging and stereotactic techniques allow physiologists and neurosurgeons to identify anatomic targets (contralateral to the affected limb) accurately, resulting in enhanced efficacy and reduced surgical morbidity. The two proved techniques are stereotactic thalamotomy and chronic thalamic deep brain stimulation (DBS), with DBS preferred.[76] Surgery is contraindicated in patients who are poor candidates due to underlying medical conditions and those with marked cognitive problems. Both procedures, when successful, allow patients to become drug free.

Stereotactic Thalamotomy. Stereotactic thalamotomy for essential tremor is an ablative technique (thermocoagulation) using a Cartesian coordinate system to target the ventralis intermedius (Vim) nucleus, a group of neurons located within the thalamus. Theories suggest that creation of a lesion in the Vim nucleus disrupts abnormal tremorogenic activity in the cerebellar-thalamic circuitry.

Most thalamotomies are unilateral, performed on the side of the brain contralateral to the dominant or most severely affected limb. Bilateral thalamotomies generally are avoided due to high risk of severe, permanent dysarthria or even complete mutism. The procedure takes 2-3 hours and is performed under local anesthesia to allow the patient to participate in physiologic examinations. Because brain tissue is devoid of pain sensation, the procedure is relatively painless. The efficacy of unilateral thalamotomy is high, with greater than 80% of patients experiencing long-lasting and complete (or near complete) suppression of the targeted tremor.[76] Often concomitant midline tremors (head, voice) also improve. If the procedure results in incomplete lesioning, the patient may experience a mild residual tremor or eventually reemergence of the targeted tremor. In such cases, repeating the operation may be an option.

With an experienced neurosurgical team, the occurrence of severe, persistent morbidity (dysarthria, contralateral motor weakness, gait abnormality, paresthesia, cognitive deficits) associated with unilateral thalamotomy is less than 2% and the risk of operative mortality is almost nonexistent.

Thalamic Deep Brain Stimulation. Unilateral and bilateral DBS is effective for patients with disabling tremor that is not adequately controlled by pharmacotherapy. It has gained wide acceptance and is preferred over thalamotomy due to advantages related to its nonablative and adjustable nature. These include reversibility due to minimal lesioning of the Vim and the ability to change impulse variables to minimize side effects and increase efficacy. An additional benefit of thalamic DBS is the ability to perform bilateral procedures with a reduced risk of permanent morbidity.

The specific mechanism of action of DBS in essential tremor remains unknown but it may suppress tremor by providing chronic artificial "neural noise" that essentially disrupts cyclic activity within the motor circuit pathway.[77] The efficacy of the procedure is at least equivalent to that of thalamotomy, and significant improvements in disability and health-related quality of life can be expected.[78] In a randomized, controlled study of 68 patients with drug-resistant tremor (45 with parkinsonism, 13 with essential tremor, 10 with multiple sclerosis), DBS was more effective than thalamotomy in improving functional ability in patients with essential tremor and parkinsonism.[76] However, after 2 years, the benefits in some patients with essential tremor appeared to have waned. Midline symptoms such as voice and head tremor were also improved, although less predictably than hand tremor. Excluding one fatality due to intracerebral hemorrhage, DBS was safe, with few patients experiencing cognitive decline, dysarthria, facial paresis, gait imbalance, or arm ataxia.

The disadvantages of thalamic DBS include increased expense over thalamotomy (mainly due to the cost of the hardware and programming instruments), the need for labor-intensive follow-up and monitoring, and the potential risk of inflammatory responses and infection due to implantation of foreign material. Batteries for the implanted pulse generator (IPG) require replacement, and the IPG or lead wires may have to be replaced due to malfunction or breakage. For patients living in rural areas where mandatory follow-up for IPG adjustments are not feasible, DBS maintenance can be problematic.