Key points

Introduction

Gastroparesis is a syndrome characterized by delayed gastric emptying (GE) and symptoms thereof in the absence of gastric outlet obstruction. In diabetes mellitus (DM), delayed GE is often asymptomatic. This term should be reserved for patients with delayed GE and upper gastrointestinal symptoms. DM and idiopathic disease are the 2 primary causes associated with gastroparesis. Only 2 studies have evaluated the epidemiology of gastroparesis in the population. Even these studies are based on data collected from patients who presented for medical attention rather from than a random sample of people in the community. Currently, GE can only be assessed with scintigraphy, which requires specialized laboratories and radiation exposure, limiting population-based studies of the epidemiology of gastroparesis. Hence, understanding of many facets of the epidemiology of gastroparesis is primarily based on case series or hospital-based databases rather than on the population. These studies suggest that gastroparesis is not uncommon and can impair quality of life. The incidence of hospitalizations associated with a diagnosis of gastroparesis has increased considerably since 2000.

Several organic diseases affect gastric neuromuscular functions through causing an extrinsic or enteric neuropathy or a myopathy. Among patients who do not have an underlying disorder that is known to be associated with gastroparesis, the pathogenesis of gastroparesis is poorly understood.

Introduction

Gastroparesis is a syndrome characterized by delayed gastric emptying (GE) and symptoms thereof in the absence of gastric outlet obstruction. In diabetes mellitus (DM), delayed GE is often asymptomatic. This term should be reserved for patients with delayed GE and upper gastrointestinal symptoms. DM and idiopathic disease are the 2 primary causes associated with gastroparesis. Only 2 studies have evaluated the epidemiology of gastroparesis in the population. Even these studies are based on data collected from patients who presented for medical attention rather from than a random sample of people in the community. Currently, GE can only be assessed with scintigraphy, which requires specialized laboratories and radiation exposure, limiting population-based studies of the epidemiology of gastroparesis. Hence, understanding of many facets of the epidemiology of gastroparesis is primarily based on case series or hospital-based databases rather than on the population. These studies suggest that gastroparesis is not uncommon and can impair quality of life. The incidence of hospitalizations associated with a diagnosis of gastroparesis has increased considerably since 2000.

Several organic diseases affect gastric neuromuscular functions through causing an extrinsic or enteric neuropathy or a myopathy. Among patients who do not have an underlying disorder that is known to be associated with gastroparesis, the pathogenesis of gastroparesis is poorly understood.

Prevalence and incidence

Only 1 study on the epidemiology of idiopathic gastroparesis in the population has been published. That study, which was conducted in the Rochester Epidemiology Project, defined gastroparesis as definite (ie, delayed GE according to standard scintigraphy and typical symptoms for >3 months), probable (ie, typical symptoms and food retention on endoscopy or upper gastrointestinal study), and possible (ie, typical symptoms alone or delayed GE according to scintigraphy without gastrointestinal symptoms). A total of 83 patients had definite gastroparesis, 127 had definite or probable, and 222 had any of the 3 definitions. On January 1, 2007, the age-adjusted prevalence of definite gastroparesis per 100,000 persons was approximately 4-fold higher in women (37.8; 95% CI, 23.3–52.4) than in men (9.6; 95% CI, 1.8–17.4). Likewise, the age-adjusted incidence per 100,000 person-years of definite gastroparesis for 1996 through 2006 was approximately 4-fold higher in women (9.8; 95% CI, 7.5–12.1) than in men (2.4; 95% CI, 1.2–3.8).

Earlier reports from tertiary referral centers observed that up to 60% of patients with long-standing type 1 DM (T1DM) and gastrointestinal symptoms had diabetic gastroparesis. However, these studies predated the routine use of intensive insulin therapy for T1DM.

More recently, population-based studies of gastrointestinal symptoms in DM have been based on symptoms alone or symptoms and delayed GE. Compared with the studies in selected populations mentioned previously, the cumulative incidence of diabetic gastroparesis among patients with DM in the community is lower. In the only community-based study from Olmsted County, MN, the cumulative incidence of symptoms and delayed GE over 10 years was 5% in T1DM (hazard ratio [HR], 33; 95% CI, 4.0, 274; adjusted for age and gender vs controls), 1% in type 2 DM (T2DM) (HR, 7.5; 95% CI, 0.8, 68; adjusted for age and gender vs controls) and 1% in controls ( Table 1 ). The risk of gastroparesis in T1DM was significantly greater than in T2DM (HR, 4.4; 95% CI, 1.1, 17). Gastroparesis was documented by physician diagnosis, evaluating GE with scintigraphy, or symptoms and retained food at endoscopy. Because gastroparesis was identified only in people who presented for care, people in whom GE was not evaluated may not have been identified. Hence, this study assessed the cumulative incidence of diabetic gastroparesis (over 10 years) rather than the prevalence of diabetic gastroparesis.

Several studies have evaluated the epidemiology of upper gastrointestinal symptoms but not GE among people with diabetes in the community. In most community-based studies, the prevalence of gastrointestinal symptoms was not significantly higher in people with diabetes than in asymptomatic controls. In the Rochester Diabetic Neuropathy Study from Olmsted County, MN, only 1% of patients had symptoms of gastroparesis ( Table 2 ). Another study from Olmsted County observed that the prevalence for nausea and/or vomiting or dyspepsia was not significantly different in people with T1DM or T2DM relative to controls. That study did not assess GE. However, in an Australian community-based study of 423 patients with predominantly (95%) T2DM, the prevalence of several upper gastrointestinal symptoms, including abdominal pain or discomfort, early satiety, postprandial fullness, bloating, heartburn, nausea, vomiting, and dysphagia, was higher in a sample of people predominantly with T2DM than in controls. Taken together, these data suggest that gastrointestinal symptoms are not uncommon among people with DM. The risk of gastroparesis is higher in T1DM than in T2DM.

Longitudinal studies from Australia suggest that, similar to functional gastrointestinal disorders, gastrointestinal symptom turnover also occurs in DM. Turnover refers to appearance and disappearance of symptoms over time. In the first study, considerable turnover in gastrointestinal symptoms was seen 3 years after the initial assessment in patients with T2DM. However, appearance and resolution were balanced; hence, overall prevalence was comparable at follow-up. Several factors, but not glycemic control, predicted symptom change; these factors varied among symptoms. In another cohort of 139 subjects with diabetes, of whom approximately 50% had T1DM and 5% had severe autonomic dysfunction, symptom turnover varied between 15% and 25% in the group with diabetes and was not significantly different from that of controls. Symptom turnover was not associated with glycemic control or autonomic neuropathy, but rather with depression (ie, appearance and disappearance of depression were associated with gain and loss of gastrointestinal symptoms, respectively).

No studies that have assessed symptoms and GE among randomly selected people in a community, perhaps because scintigraphy is cumbersome and involves radiation exposure. GE breath tests (GEBTs) use ¹³ C, which is a stable isotope, and offer an alternative approach for measuring solid-phase GE without elaborate detection equipment or radiation exposure in the office or bedside. The U.S. Food and Drug Administration (FDA) is currently reviewing a ¹³ C-spirulina GEBT for use in the United States. GE can also be measured with a nondigestible capsule, SmartPill (Given Imaging, Ltd., Yoqneam, Israel), which records luminal pH, temperature, and pressure during gastrointestinal transit, providing a measure of GE time.

Although most attention has focused on delayed GE, DM is also associated with rapid GE. In a Mayo Clinic tertiary referral study of 129 patients with diabetes and upper gastrointestinal symptoms undergoing scintigraphy, 42% had normal, 36% delayed, and 22% rapid GE. Each category had an approximately equal number of patients with T1DM and T2DM. However, the prevalence of rapid GE among people with diabetes in the community is unknown.

Hospitalization

Although the incidence of gastroparesis did not change significantly between the periods 1996 to 2000 and 2001 to 2006, data from the Nationwide Inpatient Sample (NIS), which comprises a nationally representative sample of 5 to 8 million hospitalizations, demonstrate a 138% increase in the hospitalizations related to gastroparesis between 1995 and 2004. DM was listed as a comorbidity in 21.0% of these patients in 1995 and 26.7% in 2004. Gastroparesis was the primary or secondary diagnosis for these hospitalizations. The former increased by 158% between 1995 and 2004, and most of this change (138%) occurred between 2000 and 2004. In comparison, changes were smaller over that period in diabetes-related hospitalizations (53% increase), all hospitalizations (13% increase), and hospitalizations with 4 other gastrointestinal conditions, such as gastroesophageal reflux disease (GERD), gastric ulcer, gastritis, or nonspecific nausea/vomiting as the primary diagnosis (ranging from a 3% decrease to 76% increase). Conceivably, this increase in hospitalizations for gastroparesis may reflect an increased prevalence of DM or gastroparesis; changes in diagnostic criteria, severity, and treatment of gastroparesis; or better recognition and/or diagnosis of this disorder. Several possible explanations exist for the striking increase after 2000. The withdrawal of the prokinetic agent cisapride from the market might have increased the number of symptomatic patients. The FDA provided humanitarian device exemption to gastric electric stimulation around 2000. This device often entails hospitalization for surgical placement of the stimulator. Increased recognition of the disorder gastroparesis and a change in hospital coding practices may also contribute.

Compared with the 4 gastrointestinal conditions as the primary diagnosis (GERD, gastric ulcer, gastritis, and nonspecific nausea/vomiting), the length of stay was also longer for gastroparesis listed as the primary diagnosis (increase of 15.4%–66.2%; all P <.001). The total charges for hospitalization in patients with gastroparesis listed as the primary ($20,573) or secondary ($24,965) diagnosis in 2004 were also higher than these other diagnoses except for gastric ulcer, for which the total charge was $23,259. Similar trends for gastroparesis versus other diagnoses were observed in 1995.

The Nationwide Inpatient Sample study did not clearly discriminate between hospitalization for idiopathic versus diabetic gastroparesis. In the National Institute of Diabetes, Digestive and Kidney Diseases (NIDDK) consortium, patients with gastroparesis from T1DM reported more hospitalizations (ie, 5.1 ± 6.4 per year, mean ± SD), mostly for vomiting and dehydration, than those with idiopathic gastroparesis (1.6 ± 3.0) or T2DM (2.7 ± 5.7). In a study from the Beth Israel Deaconess Medical Center and the Joslin Diabetes Center, patients with diabetic gastroparesis also had more hospital days than those with DM and gastrointestinal symptoms but normal GE (ie, 25.5 vs 5.1 per 1000 patient days).