Scott Eveloff, MD
Scott Eveloff, MD

What happens when treating one sleep disorder actually causes the worsening of another? The case of J.A., a patient who was treated for RLS only to experience a worsening sleep-related eating disorder (SRED), shows the difficulty of finding the right pharmacological solution to multiple sleep disorders.


SRED is an unusual parasomnia involving involuntary or irresistible episodes of eating and drinking arising out of sleep, ranging from full to partial to absent recall or awareness. SRED needs to be differentiated from nocturnal eating syndrome (NES), which is a disorder of excessive or compulsive eating with consumption of 25% to 35% daily caloric intake after the evening meal. In NES, eating is done with full awareness and recall, and food materials are excessive, but not bizarre.

While SRED can be idiopathic, its causes often vary. When secondary, it may be associated with primary disorders of sleep such as OSA or periodic limb movements, or restless legs. Prior sleepwalking behavior may be present in up to two thirds of afflicted patients. An additional important secondary cause can be medications, the most notorious of which is zolpidem (Ambien). Triazolam, anticholinergics, lithium, and other medications also have been implicated. Finally, SRED may be associated with stress or alcohol/substance withdrawal.


SRED patients are typically, but not exclusively, female. Classic sleep-related eating behavior involves not only the atypical timing of eating, but such behavior may also be associated with full, partial, or absent awareness and subsequent recall.

An additional hallmark of SRED is markedly excessive, at times “out of control,” caloric intake. High-calorie fats or carbohydrates are typical consumption items; healthy food items—such as vegetables or fruit—are unusual. Bizarre types of foods are also common consumption items for SRED patients. These may range from odd food combinations to eating substances normally not regarded as nutritional—such as cigarette butts, cooking grease, or frozen or uncooked foods. Behaviors may occur out of any stage of sleep in contrast to the slow-wave origin of classic sleepwalking.


J.A. is a 40-year-old woman who was initially seen for long-standing restless legs symptomatology, which started during her first pregnancy. The symptoms involved both lower extremities and one upper extremity at the time of initial evaluation.

The patient also complained of sleep eating, typically involving not only ingestion of excessive amounts of food, but food with high caloric content and/or food types she would have otherwise not eaten. Examples include green olives with mayonnaise on bread and peanut butter mixed with oatmeal. Such episodes would occur on awakening from sleep as well as when she was unable to initially fall asleep. She described these episodes as resulting from an irresistible urge to eat if she had not yet been asleep, and episodes of eating with only partial recollection or awareness when arising from sleep.

She was not taking any prescription or illicit medications, nor was there an admitted history of alcoholism, sudden alcohol abstinence, or psychiatric disorders. The patient also denied a history of childhood sleepwalking or other sleep disorders. There was no known family history of parasomnias, although she claimed her mother also suffers from restless legs symptomatology. There was no admitted personal history of illicit substance abuse or intoxication/withdrawal syndrome.

The patient’s iron studies were checked, and no iron deficiency was found. A sleep study showed no sleep apnea, no periodic limb movements, normal sleep latencies, normal sleep architecture, and no abnormalities.


The patient was started on ropinirole (Requip) empirically for the RLS. She noted significant improvement at night, but experienced persistent RLS during the day.

Her sleep eating, however, actually worsened. She noted increasingly frequent episodes, with increasing amounts of food and calories ingested. These episodes still occurred at times without recollection, and other times during wakefulness but with a virtual compulsion to eat.

Due to the worsening sleep eating behavior, the patient was started on topiramate (Topamax) at 25 mg qhs. She noted immediate and significant improvement in sleep eating. After an increase to 50 mg, the sleep eating resolved totally.

Unfortunately, despite the significant and remarkable response, several months after initiation of topiramate, the patient noted difficulty finding words and reduced clarity of thinking. Typical side effects of topiramate include peripheral paresthesias and, less commonly, blepharospasm; sedation may become evident but is not typically problematic due to bedtime dosing. Language difficulties, as classically demonstrated by the patient, can actually be an underappreciated side effect, sometimes disabling enough to require discontinuation. These language difficulties experienced by the patient persisted despite a reduction back to 25 mg per night of topiramate.

After total topiramate discontinuation, the patient noted resolution of the language difficulties. The episodes of sleep eating, however, resumed and at times occurred up to two or three times per night.

A previous dopamine agonist (ropinirole) had shown minimal effectiveness in the patient, and an alternative therapy (zonisamide—Zonegran) also may be associated with similar language difficulties. The patient had previously been treated with diethylpropion and had noted reduction in sleep eating at that time. Diethylpropion is chemically related to bupropion (Wellbutrin), which has been used anecdotally as treatment for SRED. Based on such limited experience, the patient’s own favorable experience, and bupropion’s dopaminergic activity, the patient was started on bupropion for adjunctive treatment of sleep-related eating disorder.


Currently, pharmacologic therapy for SRED rests on relatively small, uncontrolled medication trials and, at times, anecdotal experience. Future investigations of the neurohumoral pathophysiology of sleep-related eating disorders will contribute to finding alternative therapies when others either do not work or entail unacceptable adverse effects. The empiric use and possible benefit of bupropion in this patient, with its dopaminergic activity, may be one such example.

SRED should be regarded as separate from typical disorders of arousal resulting in sleep terrors, sleepwalking, and confusional arousals. Such classification is not only of academic interest, but may have implications for treatment options as well. Disorders of arousal are much more prevalent in childhood, whereas sleep-related eating disorder is typically a disease of post-pubertal/adult females. Disorders of arousal typically arise out of slow-wave (Stage 3-4) sleep, whereas sleep eating has no specific predilection for a particular part of the sleep cycle.

Perhaps of greatest interest, SRED may have an underlying dopamine-deficient basis. Dopamine agonists may reduce food intake and induce weight loss. Sleep-related eating is commonly associated with restless legs syndrome and periodic limb movements, both disorders associated with dopaminergic dysfunction. This may explain why agents extremely useful with disorders of arousal such as clonazepam have only limited efficacy in SRED. Clonazepam does have dopaminergic activity, which may differentiate its role in SRED from that of related medications such as triazolam and zolpidem. Finally, dopaminergic agents have in the past been the mainstays of treatment for SRED. Combinations used for PLMS and RLS, including dopaminergic agents and codeine, also can be useful in SRED.

Treatment options for SRED have broadened in recent years, however. First and foremost, treatment of underlying disorders possibly promoting SRED should be undertaken. Most commonly, this involves treatment of sleep-disordered breathing or periodic limb movements. Medications should be carefully reviewed, and use of or withdrawal from alcohol should be searched for and confronted if suspected. If such associated processes exist, treatment of SRED should include treatment of those processes, as they may contribute to arousal and perpetuate sleep-related behavior.

In this case, however, treatment of the patient’s RLS actually worsened the patient’s SRED.

Want more fascinating case reports? Check Sleep Review’s online archives.


Topiramate has been found to be increasingly efficacious in treating SRED, although no large-scale studies have been performed. The mechanism of action underlying such benefit is unclear. Other treatment including topiramate has been shown to have appetite-suppressant properties in patients without SRED, and can be used as an adjunctive weight loss agent. Isolated case reports also have demonstrated the utility of bupropion, an antidepressant with dopaminergic and noradrenergic properties, as well as trazodone. Often response to a single agent is suboptimal, requiring combination therapy in patients suffering from SRED.

Scott Eveloff, MD, is assistant medical director at SomniTech Inc, Overland Park, Kan. He can be reached at .