Case Study: Chronic Vomiting in a Patient With Type 2 Diabetes

A.M. is a 54-year-old woman who was diagnosed as having type 2 diabetes 8 years ago. She has mild hypertension but no evidence of coronary artery disease or other complications of diabetes. She is managed with insulin and has poor control of hyperglycemia; recent HbA_1c concentrations range from 10 to 12%.

One year ago, the patient began having episodic nausea and vomiting. The problem increased in frequency, and for the past 6 months some degree of nausea and vomiting has been present daily. Symptoms occur most commonly in the morning and with eating, when early satiety, bloating, and fullness also are present. Metabolic regulation has worsened, although her weight has remained stable.

Physical examination was normal with the exception of mild obesity. Investigation with endoscopy, hepatobiliary ultrasound, abdominal CT scan, and routine laboratory was unrevealing. A solid-phase gastric emptying study showed mild delay with abnormal radionuclide retention at 2 h. Metoclopramide was transiently effective, and erythromycin produced no benefits.

1. Does A.M. have gastroparesis?

2. Would a more detailed neuropsychiatric evaluation be helpful?

3. What additional management options could be tried?

Chronic nausea and vomiting are particularly detrimental for diabetic patients. The symptoms not only are stressful and uncomfortable, but also interfere with daily functioning and promote deterioration of metabolic control. Vomiting is a source of high hospitalization costs and health care utilization by diabetic patients.

Gastroparesis is defined by the presence of delayed gastric emptying without mechanical obstruction. The abnormality can be demonstrated by retained food on imaging studies or endoscopy following an overnight fast or, more precisely, by delayed clearance of radionuclide on a quantitative gastric emptying study, as occurred in this case. Obstruction in the antropyloric region, duodenum, or proximal small intestine (e.g., from peptic ulcer disease or tumor) must be excluded as the cause of delayed emptying. This usually is accomplished with endoscopy, as was done for A.M., who indeed has gastroparesis.

Unfortunately, upper gastrointestinal symptoms, especially nausea and vomiting, are too frequently attributed to gastroparesis, which then becomes the focus of therapy. Gastroparesis is found in diabetic as well as nondiabetic subjects and has many causes. Delayed gastric emptying, like slow transit in other gut organs, appears to be one of the gastrointestinal sequelae of diabetes, but because of its many other causes, the relationship of gastroparesis to markers of advancing diabetes (including neuropathy) is only modest in symptomatic patients. As many as 40% of nondiabetic patients with functional nausea and vomiting and no definable pathological explanation for symptoms also have delayed gastric emptying. Thus, gastroparesis should be considered a nonspecific finding in diabetes and may be unrelated to the metabolic disorder.

The relationship of symptoms to delayed emptying is even more precarious in most cases. Gastroparesis typically is asymptomatic in diabetic patients. The poor correlation between degree of symptoms and degree of emptying delay is known, but its acknowledgment has been sluggish, even by the gastroenterological community. Results of treatment trials have further supported the dissociation.

Regular metoclopramide (Reglan) therapy can improve persistent nausea and vomiting in diabetic patients, but many clinicians are unaware that its effects on gastric emptying often are insignificant after as little as 2 months of continuous use. Likewise, the beneficial effects of domperidone (Motilium) (another dopamine-antagonist prokinetic agent that is not approved for use in the United States) on vomiting cannot be attributed to improvement in gastric emptying. Both of these agents have central antiemetic effects that appear to explain their sustained benefits in clinical practice.

No change in gastric emptying accompanies spontaneous symptom remission in most diabetic patients presenting with vomiting. Thus, the finding of gastroparesis should not restrict the evaluation for alternative causes of nausea and vomiting, nor should it necessarily be attributed to diabetes or limit interventions to approaches that accelerate gastric emptying.

Finding evidence of peripheral or autonomic neuropathy might offer support for a diabetic etiology to the delay, but it would provide little other direct clinical benefit. Other neurological symptoms and findings are usually present in patients with central nervous system (CNS) diseases that produce vomiting (e.g., malignancy, infarction, infection). Psychiatric symptoms of anxiety or depression are common in patients with functional nausea and vomiting but have the same nebulous role in the initiation or perpetuation of these symptoms as they do in other functional gastrointestinal syndromes.

Nevertheless, resolution of active psychiatric illness may be required to reduce symptoms to an acceptable level. In cross-sectional evaluation of diabetic patients with gastrointestinal symptoms (including nausea and vomiting), psychiatric symptoms predict both the presence and severity of somatic symptoms—in contrast to the poor predictive value of gastroparesis. Consequently, at least a cursory psychiatric assessment may have important treatment implications.

The focus on delayed gastric emptying has left under-managed many diabetic patients with chronic nausea and vomiting. “Prokinetic” agents typically are offered first and may be beneficial for some patients. But the benefits are unlikely restricted to prokinetic actions. As mentioned, metoclopramide and domperidone have pronounced central antiemetic effects that are sustained. Cisapride (Propulsid), now with very limited availability in the United States, was promoted for its prokinetic properties and has more durable effects on gastric emptying than other compounds. This 5HT4 agonist, however, also favorably influences visceral sensory input to the CNS and potentially through this mechanism has been effective in nondiabetic patients with functional symptoms (with or without gastroparesis). Erythromycin is primarily prokinetic in action, but its beneficial effects in diabetes are limited.

When prokinetic treatment trials are exhausted, clinicians often are at a loss for alternative therapies. Based on the available information, particularly the poor association of gastroparesis with symptoms, we prefer to manage diabetic patients in the same fashion that we would nondiabetic patients with chronic functional gastrointestinal complaints. In some patients, anticholinergics/antispasmodics, as contrary as it may seem, can be beneficial. Other conventional antiemetic agents (e.g., prochlorperazine [Compazine], promethazine [Phenergan]) may be useful, but, because of side effects, are less appealing for chronic, daily use.

Antidepressants, particularly the tricyclic antidepressants in low daily dosage, can be particularly helpful. These medications benefit more than 80% of nondiabetic patients with chronic vomiting syndromes who have failed other approaches, the benefits not conspicuously related to psychiatric drug actions. Preliminary data support a similar efficacy with open-label use of tricyclic antidepressants in diabetic subjects. A.M. was treated and maintained on nortriptyline (Pamelor), 50 mg/day at bedtime, with prompt and sustained response.

Contemporary antidepressants, including SSRIs, may work through a different mechanism and have been less studied for this indication. We reserve them for patients with active anxiety and/or depression symptoms that may be interfering with treatment response.

Psychotherapy and behavioral interventions also may have a role in refractory cases. Implanted electrical gastric stimulators, initially investigated as potential prokinetic devices, are now being explored actively in diabetic and nondiabetic subjects with chronic nausea and vomiting for their beneficial effects on visceral afferent (sensory) modulation rather than gastric emptying.

• Gastroparesis, or nonobstructive delayed gastric emptying, should not be equated unconditionally with upper gastrointestinal symptoms.

• Most available treatments for nausea and vomiting in diabetic patients do not exert their dominant effects through improving gastric emptying, at least when used chronically.

• The range of treatments for functional nausea and vomiting in nondiabetic patients, including tricyclic antidepressants, also should be considered in patients with diabetes and chronic symptoms.