Body weight in patients with idiopathic gastroparesis.

The NIDDK Gastroparesis Clinical Research Consortium (GpCRC) is a network of seven clinical centers and Scientific Data Research Center. Patients were enrolled in Gastroparesis Registry 2 (ClinicalTrials.gov Identifier: NCT01696747 ) which was an observational study of patients with gastroparesis ( 7 ) with entry criteria being 18 years or older with symptoms of at least 12 weeks, gastric emptying scintigraphy (GES) within 6 months of enrollment, and no structural abnormality as seen by upper endoscopy within one year of enrollment. The patients enrolled in the registry could have either idiopathic, diabetic, or postfundoplication gastroparesis. At each study visit (enrollment, 6 months, and 12 months), symptoms and treatments were assessed, questionnaires were filled out, and body weight and height were obtained. Patients received standard of care evaluations during follow-up, allowing adjustments or additions of treatments for their clinical care. This study reports on the idiopathic gastroparesis patients who had one year follow up information in order to look at changes in body weight over time. Idiopathic etiology was based on no previous gastric surgery, no diabetes history, and no other known etiologies. The study was approved by the Institutional Review Board at each center. All authors had access to study data and reviewed and approved the final manuscript. During face-to-face interviews with each patient, study physicians or coordinators completed case report forms including data on gastroparesis disease onset, symptoms, body weight, disease profile, associated medical conditions including diabetes, and medication and supplemental therapies. Prior surgeries were inquired. Study physicians performed a physical examination including body weight and height. Laboratory measures included complete blood count (CBC), hemoglobin A1 C , and inflammatory markers {erythrocyte sedimentation rate (ESR), and C-reactive protein (CRP)}. The clinical severity of gastroparesis was ranked as grade 1: mild gastroparesis (mild symptoms relatively controlled); grade 2: compensated gastroparesis (symptoms only partially controlled with medications); grade 3: refractory gastroparesis (refractory symptoms that are not controlled) ( 8 ). Each patient filled out the Patient Assessment of Upper GI Symptoms (PAGI-SYM) questionnaire which assesses symptoms of gastroparesis, dyspepsia, and gastroesophageal reflux disease ( 9 ); it includes the nine symptoms of the Gastroparesis Cardinal Symptom Index (GCSI) ( 10 ). Patients were asked to assess the severity of their symptoms during the previous two weeks from 0 to 5 where no symptoms=0, very mild=1, mild=2, moderate=3, severe=4, and very severe=5. PAGI-SYM subscale scores for upper GI symptoms were calculated for nausea/vomiting, postprandial fullness/early satiety, bloating, upper abdominal pain/discomfort and heartburn/regurgitation. Overall gastroparesis symptom severity was determined by the GCSI score, which averages the three subscores of nausea/vomiting, postprandial fullness/early satiety, and bloating subscores. The nausea/vomiting subscale score was the mean of scores for nausea, retching, and vomiting. The postprandial fullness/early satiety subscale score was the mean of scores for stomach fullness, not able to finish a normal-sized meal, feeling excessively full after meals, and loss of appetite. The bloating/visible distention subscale score was the mean of scores for bloating and stomach or belly visibly larger. Quality of life was assessed with two validated instruments. The Medical Outcomes Study 36-Item Short-Form Health Survey version 2 (SF-36v2) was used to assess the patients’ overall physical and mental health ( 11 ). Disease-specific quality of life was assessed by Patient Assessment of Upper Gastrointestinal Disorders Quality of Life (PAGI-QOL) survey, which scores 30 factors from 0 (none of the time) to 5 (all of the time) ( 12 ). Patients were asked how often their gastrointestinal problems have affected different aspects of their quality of life and well-being in the past two weeks. The factors were combined into five subscales: daily activities, clothing, diet, relationship, and psychological wellbeing & distress subscales. Overall PAGI-QOL scores were calculated by taking means of all subscales after reversing item scores; low scores reflect poor quality of life while higher reflect better life quality. Psychological functioning was assessed using Beck Depression Inventory (BDI) and State-Trait Anxiety Inventory (STAI). BDI is a 21-question inventory assessing depression, cognition, and physical well-being ( 13 ). Each answer is scored on a scale of 0 to 3. Higher total scores indicate more severe depressive symptoms. STAI consists of 20 questions relating to state anxiety (a temporary or emotional state) and 20 questions pertaining to trait anxiety (long standing personality trait anxiety with a general propensity to be anxious) ( 14 ). We assessed oral caloric consumption and energy expenditure of the patients. Block 2005 Food Frequency Questionnaire is a 110 food item questionnaire that estimates usual intake of nutrients and food groups ( 15 ). Block Energy Expenditure Survey - Adults was used to measure total average energy expenditure per day ( 16 ). Patients stopped medications that could affect gastrointestinal motility for 72 hours prior to the study and reported in the morning after fasting overnight. Diabetic patients had glucose checked to ensure 60 % at 2 hours and/or >10% at 4 hours was considered delayed gastric emptying. Patients stopped medications that could affect gastrointestinal motility for 72 hours prior to the study and reported in the morning after fasting overnight. Diabetic patients had glucose checked to ensure <275 mg/dl. Subjects ingested water until they achieved the sensation of “completely full” during a five-minute time period ( 19 ). The volume of water ingested was recorded. Baseline patient characteristics were compared by body weight categories (by BMI in kg/m 2 : underweight or normal, BMI < 25; overweight, 25 ≤ BMI < 30; obese, BMI ≥ 30). Data are presented as means±SD or N (%) and p-values determined by Cochran’s chi-square test for trend for categorical measures and by linear regression with ordinal predictor for continuous measures. Baseline characteristics included age, gender, ethnicity, race, diabetes status, gastroparesis etiology, acute vs. insidious symptom onset, presence of initial infectious prodrome, hospitalization for gastroparesis in the year prior to enrollment, 2-hour and 4-hour gastric emptying retention (%), GCSI, PAGI-QOL, SF-36 physical and mental scores. Multiple linear regression analyses using Akaike Information Criteria (AIC) with forward selection were performed to determine a reduced set of factors associated with body weight. AIC estimates the relative amount of information lost by a given model, with less information lost indicating a better model ( 20 ). We also performed similar analyses looking at the relationships of 48-week change of body weight with symptoms at enrollment as well as change of body weight with 48-week change in gastroparesis symptoms. All p-values are two-sided and not corrected for multiple comparisons since these are exploratory analyses. Analyses were conducted using SAS 9.4 ( 21 ) and Stata 15.1 ( 22 ).

The 138 patients with idiopathic gastroparesis who were followed for at least 48 weeks in NIDDK/NIH Gastroparesis Registry 2 made up the cohort for this study. Among the 138 participating patients, 128 were female (93%), 130 (94%) were white, with average age of 42 years. 18 patients had a history of connective tissue disorders (defined as scleroderma, lupus, Raynaud’s phenomenon or other autoimmune disorder). Figure 1 shows a histogram depicting the distribution of BMI for the 138 patients with idiopathic gastroparesis. At time of enrollment, of the 138 patients with idiopathic gastroparesis followed in the registry, 14 [10.1%] patients were underweight (BMI<18.5), 54 [39.1%] were normal weight (BMI 18.5 to 25), 30 [21.7%] were overweight (BMI 25 to 30), 18 (13.0%) had class 1 obesity (BMI 30 to 35), 13 (9.4%) had class 2 obesity (BMI 35 to 40) and 9 [6.5%] were class 3 obesity (BMI ≥40). Supplementary Table 1 shows the Mean (SD) of GES at 4 hours and total GCSI by these BMI categories For analysis, patients were divided into three categories: underweight/normal weight (BMI<25); overweight (BMI 25 – 29.9); and obese (BMI ≥30). The 14 patients in the underweight group were too few to be analyzed as an individual group. Table 1 shows the baseline characteristics of the patients with idiopathic gastroparesis stratified by BMI. Overweight/Obese idiopathic gastroparesis patients had an increased Hispanic ethnicity (p=0.05). We have previously reported that a higher proportion of Hispanics with gastroparesis have diabetic etiology than non-Hispanic whites ( 23 ). In the 6 months prior to enrollment, 38 (27.5% of patients) had been losing weight, 66 (47.8% of patients) had been staying at the same weight, whereas 34 (24.6% of patients) had been gaining weight. These percentages among the underweight/normal, overweight, and obese groups were not statistically different (p=0.10), although the underweight/normal group had gained more weight in the 6 months prior to enrollment (5.9 lbs) than the other body weight groups (p=0.05). Specifically, of the 40 obese patients, 28% had lost weight over the 6 months prior to enrollment whereas 28% had gained weight over the prior 6 months prior to enrollment. Female overweight/obese patients tended to have more polycystic ovarian syndrome (PCOS) (p=0.004). Obese patients had increased prevalence of thyroid disease, though treated, as TSH was normal and not different among the weight groups. Medications used in the 6 months prior to enrollment were ascertained. Corticosteroids were used more often in the obese gastroparesis patients (p=0.03). There were no significant differences in the use of narcotic pain medications (p=0.10), neuropathic pain medications (p=0.55) or antidepressants (p=0.40). As compared to non-obese patients, obese gastroparesis patients had less severity of early satiety (p=0.06), but not other symptoms and did not differ with respect to longer duration of gastroparesis symptoms (p=0.10). The QOL was not different among the groups as assessed with PAGI-QOL items, SF-36 physical or mental component scores. There were no differences in Beck Depression Index or State/Trait Anxiety scores. Obese patients did not differ from non-obese idiopathic gastroparesis patients in severity of gastric retention at 2 hours (p=0.21), or 4 hours (p=0.73) ( Table 1 ). Although overweight and obese patients with idiopathic gastroparesis were able to drink more during the water load test, this did not reach statistical significance (385 ml for obese, 414 for overweight, and 328 for under/normal weight; p=0.13) ( Table 1 ). The obese gastroparesis patients had increased inflammatory markers compared to nonobese patients: ESR (22 mm/hr; p=0.004), CRP (4.6 mg/dl; p=0.02), and WBC count (8100 cells/μL; p=0.0001). In these idiopathic gastroparesis patients, HgbA1c was slightly greater in the obese patients than in the nonobese patients (5.7%; p<0.0001). Using the Block Food Frequency Questionnaire, patients reported consuming similar daily calories across 3 BMI categories (p=0.42). The Block 2005 FFQ captures use of sodas (regular or diet). Overall 30% of our patients were consuming >=1 soft drink per week and 19% were consuming >2 soft drinks per week. Increasing BMI was related to increasing prevalence of consuming one or more soft drinks per week (p=0.05) ( Table 1 ). Using the Block Energy Expenditure Survey, patients that were obese reported more energy expenditure compared to normal or underweight patients (p<0.0001). There was not a significant increase of nutrition consults among the different weight categories (p=0.36) or patients stating that they were following a gastroparesis diet (p=0.54). Multiple linear regression of enrollment characteristics on absolute body mass index at baseline was performed ( Table 2 ). Body weight was positively associated with caloric consumption (p<0.001), nutrition consultation after onset of gastroparesis (p=0.001), following gastroparesis diet (p=0.04), upper abdominal pain score (p=0.01), HbA1c (p=0.008), WBC counts (p=0.002), use of J-tube (p=0.08), and inversely associated with energy expenditure (p=0.05), alcohol drinks >2 per month (p=0.003), bloating subscale (p<0.001), current smoking (p=0.03), PAGI-QOL clothing subscale (p=0.04). While in the registry, patients received standard of care during follow-up, allowing treatments for their clinical care, which could include nutritional, pharmacological, and surgical treatments. Only 2 out of 133 patients with idiopathic gastroparesis at enrollment developed diabetes at 48 weeks. Overall in the total 125 patients with PAGI-SYM data at enrollment and 48 weeks, the mean symptom severity decreased slightly over 48 weeks. The GCSI represents symptom severity from 0 (none) to 5 (very severe). The total GCSI decreased over 48 weeks by average of 0.3 (95% CI = 0.1, 0.5; p<0.001) with decreases in nausea/vomiting subscale of 0.5 (95% CI = 0.3, 0.8; p<0.001), postprandial fullness/early satiety subscale of 0.4 (95% CI = 0.2, 0.6; p<0.001) but not bloating subscale (decrease of only 0.1 (95% CI=−0.2, 0.3; p=0.67). There were also mean decreases in upper abdominal pain subscale of 0.4 (95% CI = 0.1, 0.7; p=0.003) and heartburn/regurgitation subscale of 0.3 (95% CI = 0.1, 0.4; p=0.004). Table 3 shows the change in symptoms over 48 weeks by enrollment body weight. There were no significant differences in the gastroparesis symptom 48-week changes with GCSI decreasing by 0.4 in underweight/normal weight patients, 0.1 in overweight patients, and 0.3 in obese patients (p=0.81). The heartburn/regurgitation subscale decreased the most in the obese patients [0.0 in underweight/normal weight patients, 0.4 in overweight patients, and 0.5 in obese patients (p=0.02)]. The gastric emptying improved slightly over the 48 weeks with a decrease in the 4 hour percent retention of 10% (95% CI = 6, 14; p<0.001) in the 110 patients with both enrollment and 48-week scintigraphy. This reduction was not different among the three BMI categories (p=0.50) ( Table 3 ). Of the 138 patients, 24 (18%) lost weight (>5% BW), 73 (53%) remained the same weight (within 5% BW) and 41 patients (30%) gained weight (>5% BW) over the 48 weeks after enrollment ( Table 4 ). Change in weight was not related to weight on enrollment (p=0.28). Patients gaining weight used less narcotic pain medications on enrollment (p=0.02), but had similar utilization of neuropathic pain medications (p=0.69), antidepressants (p=0.70), and corticosteroids (p=0.34). Interestingly, there was greater gastric retention at enrollment for patients gaining weight and less gastric retention for patients losing weight compared to patients that had no change in weight (p=0.04 for 2 hour solid retention and p=0.05 for 4 hour gastric retention). Patients gaining weight consumed more calories than those patients losing weight (p=0.05), but there was no difference in energy expenditure. There were no significant differences on the severity of gastroparesis symptoms at enrollment of the patients losing, staying the same, or gaining weight (p=0.67). There was an increase in severity of constipation with increasing body weight (p=0.04). There were no differences in baseline QOL, Beck Depression Index, and State/Trait Anxiety scores among the different weight change groups. Multiple linear regression of the absolute percent change in body weight over 48 weeks on enrollment characteristics was performed ( Table 5 ). Weight gain over 48 weeks was positively associated with energy consumption (p=0.003), constipation severity (p=0.005) and inversely associated with lower abdominal pain subscale (0.007). Table 6 shows changes in weight over 48 weeks with changes in symptoms over 48 weeks. Of the patients gaining weight, there was a significant reduction in the GCSI total score compared to those losing weight or staying the same (p=0.02). This was associated with a significant decrease in the postprandial fullness/early satiety subscale (p=0.0006) andsd the upper abdominal pain subscale (p=0.01). Figure 2 shows the change in the PP/ES subscale with body weight. Of the individual symptoms of the postprandial/early satiety subscore, each of individual symptoms decreased over time more in the patients gaining weight – stomach fullness (p=0.02), early satiety (p=0.003), postprandial fullness (p=0.03), and loss of appetite (p=0.003). The change in body weight was not associated with significant changes in medications during the 48 week follow up period. There were no significant changes in the PAGI-QOL total score, although there was a significant improvement in the diet subscale in the patients gaining weight (p=0.03). Multiple linear regression of absolute percent change in body weight over 48 weeks on changes in symptoms at 48 weeks was performed ( Table 7 ). Change in weight over 48 weeks was inversely associated with early satiety severity (p<0.001).

This study investigated factors that may impact body weight and changes in body weight in patients with idiopathic gastroparesis. Being underweight occurred in 10% of patients whereas obesity occurred in 29% of patients with idiopathic gastroparesis. Body weight at enrollment was associated with oral caloric consumption, energy expenditure, following a gastroparesis diet, nutrition consultation, and symptoms of abdominal pain and bloating. Weight gain over the 48 weeks follow up was positively associated with improvement in inability to finish meal, oral caloric consumption and constipation severity, and negatively associated with lower abdominal pain severity. Thus, in patients with idiopathic gastroparesis, diet, activity, and symptoms are important factors associated with body weight. In this registry study, underweight (BMI <18.5) was present in 10% of idiopathic gastroparesis patients. Early satiety, nausea, and vomiting can lead to poor oral intake and weight loss. Low body weight and decreasing body weight suggest the possibility of malnutrition and need for aggressive treatment which may include enteral feedings ( 1 ). A prior report of our consortium showed that nutrition consultation increased the chances that daily total energy requirements were being met ( 24 ). Obesity was present in 29% of patients with idiopathic gastroparesis in the cohort reported here from the second NIDDK/NIH gastroparesis registry. Interestingly, a similar number (26%) of patients were found to be obese from a prior study of ours on idiopathic gastroparesis using a different cohort of patients in our first registry ( 1 ). Overweight and obese patients have been reported by others ( 25 ). Of note, the prevalence of obesity in this series of idiopathic gastroparesis is lower than for adults in the general population (29% vs 42%) ( 26 ). The reason for obesity in patients with delayed gastric emptying has been unclear. In this study, 28% of the obese patients were actually losing weight at the time of enrollment, similar to our prior study where 30% of the overweight/obese gastroparesis patients were losing weight at the time of enrollment ( 3 ). Idiopathic gastroparesis can be a post-gastroenteritis sequelae, which can occur in obese as non-obese individuals. Medication use, lack of physical activity, use of liquid nutrient supplements, may all also play a role. The overweight patients had a higher hemoglobin A1c values than normal weight patients. Perhaps some of the patients had prediabetes; only 2 out of 133 patients developed diabetes at 48 weeks. Obese patients were more likely to be using corticosteroids and tended to use more narcotic analgesics, but not antidepressant agents or neuropathic pain medications. We did not capture the reason patients might have taken the corticosteroids, and could have been taking them anytime during the preceding 6 months to enrollment. The overweight patients had less severity for loss of appetite and inability to finish a meal compared to normal weight individuals. In this study, multiple linear regression was used to relate independent effects of specific factors to body weight. Body weight was associated with oral caloric consumption, energy expenditure, use of gastroparesis diet, and symptoms of abdominal pain and bloating. A small study, showed that patients with idiopathic gastroparesis, although in energy balance, consumed and expended less calories than healthy controls ( 27 ); a subgroup of patients with gastroparesis gained weight were less symptomatic, had increased caloric intake, and reduced energy expenditure. This present study shows that the two phenotypes of gastroparesis by BMI (overweight/obese versus normal/underweight) can be explained by a single pathophysiology- early satiety, confirmed by oral tolerance testing. Obese patients with idiopathic gastroparesis tended to have less early satiety and ingest more during the water load test. Increased consumption during the water load test suggests intact or enhanced accommodation. Greater body mass index has been associated with greater gastric volume and reduced satiation ( 28 ). Low water load ingestion has been associated with increased severity of early satiety ( 28 ). Our follow-up data provides helpful information for physicians who may want to intervene in a given patient. The symptom improvement and gastric emptying improvement over 48 weeks was not affected by the body weight status on enrollment. In patients with gastroparesis, changes in body weight, either weight loss or weight gain, often occur over time. Understanding the predictive factors for weight loss or gain can be helpful. Gaining weight over 48 weeks was associated with a significant reduction in the GCSI total score compared to those losing weight or staying the same weight. This was associated with a significant decrease in the postprandial fullness/early satiety subscale, the upper abdominal pain subscale and a trend to reduction in the nausea/vomiting subscale. Our results suggest that weight gain over 48 weeks was associated with increased oral caloric consumption and increased severity of constipation whereas weight loss was associated with increased lower abdominal pain subscale, and tended to be associated with increased retching score and a worsening in early satiety severity. Thus, weight gain is associated with a decrease in symptoms of gastroparesis and weight loss associated with an increase in symptoms, suggesting the symptoms are in part causing the weight change, rather than the weight change causing the symptoms. In this study, multiple linear regression was also used to change in body weight over time was associated with oral caloric consumption and symptoms of abdominal pain, retching, constipation and changes in symptoms of early satiety. In analogous fashion, an increase in early satiety severity may be a predictor of weight loss and therefore possible intervention. Prior studies have suggested that obese people have a faster gastric emptying than normal weight individuals ( 29 ). In our study, gastric emptying was not significantly different in the weight groups. However, there was greater gastric retention at enrollment for patients gaining weight and less gastric retention for patients losing weight compared to patients that had no change in weight. We had speculated that perhaps obesity was playing an etiological role or contributing to severity of gastroparesis based on outcomes in gastroparesis ( 30 ). Myenteric inflammation has been noted in the stomach of some patients with gastroparesis ( 31 , 32 ). Obesity was associated with higher levels of inflammatory markers, ESR and CRP. The associations of CRP, gastric retention, and water load test with three BMI categories were significant but didn’t show a clearly increasing association with increasing body weight. This observational study assessed a large number of patients with idiopathic gastroparesis in multiple centers with detailed evaluation and careful attention to body weight. This study though has some limitations. In this study, we assessed the change in weight over 48 weeks, the time period often used by nutritionists to assess body weight changes ( 1 ). Patients were treated for their gastroparesis and other disorders while in the registry according to standard of care at their center. This was not a controlled trial altering body weight; which remains to be done. We assessed dietary intake to estimate energy consumption using the Block Food Frequency questionnaire which uses dietary recall; this is helpful to assess for differences between groups of patients as was performed here, but less reliable in estimating the absolute energy intake. The Block 2005 FFQ assess soft drink consumption, but does not assess the use of liquid nutrient supplements. Patients were recruited at tertiary academic medical centers and probably reflect more symptomatic patients or participants that may have additional disorders that might affect their body weight. We did not have a large number of patients that were underweight in this study and had to combine the underweight and normal weight patients. A number of our patients were taking narcotic analgesics, which themselves can produce symptoms and delay gastric emptying. It would also be interesting to study the effects of stopping opioids on symptoms and weight. Investigating factors that might impact body weight in the community setting might also be helpful. In conclusion, this study found that underweight patients comprised 10% of our idiopathic gastroparesis cohort whereas obesity occurred in 29%. Body weight was associated with oral caloric consumption, energy expenditure, use of gastroparesis diet, nutrition consultation, and symptoms of abdominal pain and bloating. Change in body weight over time was associated with oral energy consumption and symptoms of abdominal pain, retching, constipation and changes in severity of early satiety. Thus, diet, activity, and symptoms are important factors regulating body weight in patients with idiopathic gastroparesis and comprise areas that can be addressed in managing weight issues in patients with gastroparesis.

In the clinical assessment of patients with gastroparesis, particularly their nutritional status, two factors that are often used are body weight and changes in body weight ( 1 ). The classic clinical picture of gastroparesis is a patient who is underweight and continues to lose weight due to their symptoms of early satiety, nausea and vomiting ( 2 ). However, a number of patients with gastroparesis are obese ( 3 ). What sets body weight in patients with gastroparesis is unclear. In normal individuals, body weight is thought to be related to the balance between food intake and physical activity ( 4 ). In gastroparesis, these factors, as well as symptoms of gastroparesis, medications (especially psychotropic agents, opiate medications, and corticosteroids), and inflammation may impact on body weight ( 5 ). Why patients with gastroparesis might be obese is unknown. Obesity and its associated systemic inflammation ( 6 ) could be playing a causal role for gastroparesis, obesity and gastroparesis might share overlapping etiologies, or they may simply coexist. Our aim was to investigate the factors impacting on the body weight of patients with idiopathic gastroparesis. Symptoms and other factors were investigated including food intake, physical activity, inflammatory markers, and gastric emptying. We also explored how changes in body weight over time are associated with clinical presentation as well as changes in symptoms to help determine if symptoms regulate body weight or vice versa.