Parts of this work were presented at the 20th Symposium on Neurogastroenterology and Motility, Toulouse 2005, and at the 16th International Congress on Parkinson's Disease and Related Disorders, Berlin 2005.
Predictors of gastric emptying in Parkinson's disease
Version of Record online: 10 APR 2006
Neurogastroenterology & Motility
Volume 18, Issue 5, pages 369–375, May 2006
How to Cite
Goetze, O., Nikodem, A. B., Wiezcorek, J., Banasch, M., Przuntek, H., Mueller, T., Schmidt, W. E. and Woitalla, D. (2006), Predictors of gastric emptying in Parkinson's disease. Neurogastroenterology & Motility, 18: 369–375. doi: 10.1111/j.1365-2982.2006.00780.x
- Issue online: 10 APR 2006
- Version of Record online: 10 APR 2006
- Received: 4 October 2005 Accepted for publication: 16 December 2005
- gastric emptying;
- motor impairment and disability;
- stable isotope breath tests
Abstract Predictors of gastric emptying (GE) in patients with idiopathic Parkinson's disease (PD) of a solid and liquid meal are not well defined. For measurement of GE 80 patients with PD were randomly assigned to receive either a solid meal (250 kcal) containing 13C-octanoate (n = 40) or a liquid meal (315 kcal) with 13C-acetate (n = 40). All patient groups were off medication affecting motility and were matched for age, gender, body mass index, disease duration and severity, using Unified Parkinson's Disease Rating Scale (UPDRS). Gastric emptying was compared with a healthy control group (n = 40). Multiple regression analysis was used to determine predictors of gastric emptying. Exactly 88% and 38% of PD patients had delayed GE of solids and liquids respectively. Solid and liquid emptying was similar in women and men. There were no differences in GE in PD patients <65 years of age when compared with patients ≥65 years. Multiple regression analysis showed that motor handicaps such as rigour and action tremor are independent predictors of solid GE (r = 0.68, P < 0.001). The severity of motor impairment, but not any other neurological symptom, as assessed by UPDRS is associated with gastroparesis in PD and solid emptying is more likely to be delayed.
Gastrointestinal dysfunction is a well recognized feature in patients with Parkinson's disease (PD). There is an increasing evidence indicating that in PD gastrointestinal dysfunction is caused not only by the direct involvement of the intrinsic, but also by the extrinsic, innervation of the gastrointestinal tract.1–3 Recent neuropathological findings, which are in concordance with a novel clinical staging system in PD, put emphasis on enteric dysfunction as one of the initial pathophysiological events in this disease.4,5 Gastrointestinal dysfunction can involve virtually all levels of the gastrointestinal tract. Disordered salivation, dysphagia, gastroparesis, constipation and defecatory dysfunction have been the subject of many studies.6–8 Gastroparesis, or delayed gastric emptying (GE), is reported in up to 100% of PD patients6,9 and produces symptoms such as early satiety, abdominal discomfort with bloating, nausea, vomiting, bodyweight loss and even malnutrition. Gastroparesis in PD has been assessed in only a few detailed prospective studies, but it represents a clinically relevant problem because of potential sequelae of upper gastrointestinal symptoms and profound alterations in pharmacokinetics of anti-parkinsonian drugs. Delayed drug absorption can result in erratic responses, possibly constituting one of the mechanisms for the development of motor fluctuations.6,10
Predictors of GE in patients with PD are not well defined. Although dyspeptic symptoms like bloating or nausea are reported to occur in up to 45% of the patients with PD,8 it is uncertain as to whether gastroparesis can be predicted by gastrointestinal or even neurological symptoms or demographic data. This indicates that the status of gastric motor function in PD remains to be clearly defined. Furthermore, it is unknown if delayed emptying of liquid nutrient meals occurs as frequent as delayed emptying observed after ingestion of solids by scintigraphy.10–12 As solid and liquid emptying are driven by different and distinct mechanisms, the predictors of delay in liquid emptying may differ from predictors of delayed solid GE.13
The technique most commonly used to measure GE in PD patients is scintigraphy, which requires administration of a radiolabelled test meal. This technique is accepted as the standard non-invasive test to measure GE. As scintigraphic measurements can only be accurate if movements of the patients are avoided during measurement, radioscintigraphy may not be appropriate for PD patients of advanced stages with unpredictable fluctuations or dyskinesia. In a previous study we used a stable 13C isotope breath test as a non-invasive feasible alternative method without ionizing radiation to measure GE in PD patients and showed that this method is suitable for PD patients with advanced stages.14
The aim of the present study was to investigate the prevalence of delayed GE for solids and for liquids in a series of PD patients with different disease severity and to assess possible predictors of GE of solids and liquids using stable isotope breath tests.
Materials and Methods
Written informed consent was obtained in all cases and the study was approved by the local ethics committee of Ruhr-University Bochum, Germany.
In a prospective cross-sectional study design, patients diagnosed with idiopathic PD based on the UK Parkinson's Disease Society brain bank criteria were randomly assigned to two groups to receive either a liquid test meal (group A) or a solid test meal (group B).15 Both groups were matched for gender, body mass index (BMI), disease duration and severity and other clinical parameters, as shown in Table 1. A total of 80 patients [51 men and 29 women; age: 64 ± 10 years, BMI: 25.9 ± 4.2 kg m−2 (mean ± SD)] of 140 patients screened, were included in the study. Some of the patients were included in a previous report.14 This design was chosen according to ethical considerations to prevent multiple investigations in patients with more advanced disease stages [Hoehn and Yahr (H&Y) staging scale >2.5], as all patients were examined in the ‘defined off’ state, i.e. without any anti-parkinsonian medication for at least 12 h.16
|Group A (liquid gastric emptying)||Group B (solid gastric emptying)|
|Sex (female)||17 (42.5%)||15 (37.5%)|
|Age (years)||65 ± 8||63 ± 11|
|BMI (kg m−2)||26.4 ± 3.7||25.5 ± 4.6|
|Age of symptom onset (years)||59 ± 10||56 ± 11|
|Disease duration (years)||5.8 ± 4.0||6.2 ± 4.8|
|Untreated PD||14 (35%)||8 (20%)|
|UPDRS||47 ± 20||44 ± 19|
|H&Y||2.2 ± 0.9||2.1 ± 0.9|
|Dyspeptic symptoms||15/25 (37.5%)||14/26 (35%)|
All patients were ‘off’ any anti-parkinsonian medication, i.e. 12 h, respectively, 18 h after the last standard and slow-release l-dopa tablet intake. Patients with dopaminergic drugs with a known half-life of more than 4 h were excluded from the study. In group A 14 patients were untreated (11 men and three women; age: 62 ± 8 years, BMI: 27.4 ± 4.9 kg m−2, H&Y: 1.9 ± 0.8) and in group B eight patients (three men and five women; age: 62 ± 14 years, BMI: 25.6 ± 4.1 kg m−2, H&Y: 1.6 ± 0.5). None of the subjects had a history of chronic gastrointestinal disease or gastrointestinal surgery (except for appendectomy or cholecystectomy) and none of the subjects received a medication affecting motility during the study or were taking a diet with a large amount of C4-plants like corn flour or pineapples, which have a natural 13C enrichment.17 Subjects with diabetes mellitus, thyroid disorders, autoimmune disorders, renal failure and chronic obstructive lung disease were excluded from the study. All female patients were postmenopausal and not on hormone replacement therapy.18 Before performance of the GE measurements PD patients were evaluated by a single investigator (D.W.) with version 3.0 of the Unified Parkinson's Disease Rating Scale (UPDRS)19 and with modified H&Y staging scale20,21 in the ‘defined off’ state because of interference of dopamine with GE.22,23 The investigator was blinded to the results of the GE study and gastrointestinal symptom scores. Before the GE tests, all patients were interviewed for the presence of dyspeptic symptoms defined as eight different symptoms based on the previous 2-week period (epigastric pain, bloating, postprandial fullness, early satiety, nausea, vomiting and belching).
Gastric emptying results were compared with an age- and gender-matched control range [mean ± 2 SD for liquid GE (t50: 81–131 min) and for solid GE (t50: 87–128 min)] based on results obtained from 40 healthy subjects [22 men and 18 women, mean age 61 years (range 42–89 years), mean body weight 75 kg (range 52–100 kg), mean BMI 24.7 kg m−2 (range 19.10–35.25)].
Test meals and 13C-acetate/ 13C-octanoate breath tests
After an overnight fast, at 8 am each patient of group A received 100 mg of (1-13C)-sodium-acetate dissolved in a liquid test meal (Diben®, 350 ml, 315 kcal; Fresenius Kabi GmbH, Bad Homburg v.d.H, Germany) and each patient of group B received a solid meal consisting of an egg omelette of one egg, 60 g of white bread, 5 g of margarine and 200 mL of water (53% carbohydrate, 27% lipid, 20% protein; 250 kcal) labelled with 100 mg of 13C-sodium-octanoate. Chemical and isotopic purity of both substrates were >99.5% (Euriso-top, Saint Aubin Cedex, France). Breath samples, which were exhaled into close aluminized plastic breath bags (max volume: 50 mL), were obtained before substrate administration at baseline and then in 10 min intervals for 3 h for the liquid meal and in 15 min intervals for 4 h for the solid meal. During the tests subjects remained in a relaxed sitting position. Physical activity was restricted during the test. All subjects consumed their test meals within 10 min.
The 13C/12C isotope ratio of the breath samples was analysed by non-dispersive isotope-selective infrared spectrometry (IRIS; Wagner-Analysen Technik, Bremen, Germany). The results were both expressed as delta (δ) value per mil (‰) and delta over baseline (dob = δs−δ0). The δ-value was defined as δs = (RS/RPDB−1) × 1000 [‰] with Rs = 13C/12C isotope ratio in CO2 in breath and RPDB = 0.0112372 = isotope ratio in reference (PDB = PeeDeeBelmnite, South Carolina; δs = isotope ratio in breath sample; δ0 = isotope ratio at baseline).
Analysis of 13CO2 excretion curves
To measure the metabolized proportion of 13C-sodium acetate and octanoate exhaled, the results were expressed first as a percentage dose of 13C recovered (PDR) over time for each time interval from, then the cumulative PDR (cPDR), obtained by numerical integration from PDR values, was calculated. This calculation is based on the formula proposed by Ravussin et al.24 CO2 production rate was assumed as 300 mmol per unit of body surface area per hour. Body surface area was calculated by the weight–height formula of Haycock et al.25 The percentage of 13CO2 cumulative values was fit using a model given by the formula cPDR(t) = m(1 − e−kt)β, where y is cPDR at time t in hours and m, k and β are regression estimated constants, with m being the total amount of 13CO2 when time is infinite. Gastric half-emptying time (t50) was calculated by taking PDR(t) equal to m/2 in the PDR equation, which was expressed as t50 = (−1/k)ln(1−2−1/β).26
Statistical analysis was first carried out as a descriptive evaluation of δ, PDR, cPDR, t50 and clinical characteristics of the patients (mean ± SD, unless otherwise stated). The prevalence of clinical symptoms and features (gender, untreated PD, dyspeptic symptoms) in all patient groups was compared by Fisher's exact test. Continuous data of two groups were evaluated by the Mann–Whitney U-test. Spearman rank correlation coefficients (rs) were used to compare t50 and UPDRS (total and motor subscales). A forward stepwise multiple regression analysis, based on individual t50 as the dependent parameter, was used to determine predictors of GE of solids and liquids. P < 0.05 was regarded as significant. Calculations for non-linear regression analysis and statistical analysis were made by commercial software programs (GraphPad Prism version 4.00; San Diego, CA, USA; SPSS 12.0.1; SPSS Inc., Chicago, IL, USA).
All subjects tolerated the study well, and none of the subjects experienced severe motor complications in the ‘off’ state during GE measurements.
Gastric emptying of solids and liquids, and upper gastrointestinal symptoms
After ingestion of the meals mean 13CO2 excretion in exhaled samples, reflected by t50, was different only for the solid test meal between controls and PD patients (t50 for solid meal: 167 ± 41; P < 0.001 vs controls; t50 for liquid meal: 128 ± 55; P = 0.09 vs controls; Fig. 1). Fifteen (37.5%) patients of group A had delayed emptying of liquids and 35 (87.5%) patients of group B had slow emptying of solids (Fig. 1). In nine of the 15 patients with delayed liquid emptying (60%) and in 29 of the 35 patients with delayed solid emptying (82%), GE was 3 SD outside the normal range. None of the PD patients showed an accelerated GE of liquids or solids.
Gastric emptying of liquids was delayed in three of the 14 untreated PD patients of group A (21%), whereas in all eight untreated patients of group B a delayed GE of the solid meal was observed (Fig. 1). Gastric half-emptying time was different between clinical disease severity groups for solids (H&Y 1–2.5 vs H&Y 2.5–5, UPDRS < 45 vs UPDRS ≥45, P < 0.05; Table 2). Both liquid and solid GE was similar in female and male PD patients. There were no differences in solid or liquid GE in PD patients ≤65 years of age when compared with patients >65 years (P = 0.8, Table 2). Twenty-nine (36%) patients had upper gastrointestinal symptoms, with nausea the most common of dyspeptic symptoms (in both groups 30–40%; data for other symptoms not shown). The presence of dyspeptic symptoms was not more prevalent in patients with delayed GE or in patients with higher disease severity.
|t50 (min)||Liquid gastric emptying||n||Solid gastric emptying||n|
|Women||130 ± 65||17||169 ± 50||15|
|Men||127 ± 47||23||166 ± 36||25|
|<65 years of age||124 ± 67||19||163 ± 33||22|
|≥65 years of age||131 ± 43||21||173 ± 51||18|
|UPDRS (<45)||118 ± 25||20||156 ± 30||22|
|UPDRS (≥45)||139 ± 74||20||181 ± 49*||18|
|H&Y (1–2.5)||140 ± 67||22||149 ± 28||22|
|H&Y (3–5)||113 ± 29||18||191 ± 44**||18|
Predictors of gastric emptying
There was no relationship between t50 of solids or liquids and BMI, age or disease duration. A significant correlation was observed between total UPDRS score and UPDRS part III (motor) and t50 of solids (total UPDRS: rs = 0.42, P = 0.008; UPDRS part III: rs = 0.36, P = 0.03), but not for t50 of liquids (rs = 0.2, P = 0.8). No relationship was found between UPDRS part I, II or IV score and GE.
Multiple regression analysis was used to assess predictors of t50 for liquids/solids and UPDRS part III items, which reflect the results of the neurological motor examination. In the analysis of individual scores for UPDRS part III items, item 21 for action tremor (P < 0.05) and 22 for rigor (P < 0.001) were independent predictors of solid GE. These predictors account for 46% of the variance of GE (r = 0.68, Table 3).
|Liquid gastric emptying||Solid gastric emptying|
|UPDRS 21* (action tremor)||−0.18||−0.05||0.70||−7.43||−2.42||<0.05|
|UPDRS 22* (rigour)||1.99||0.39||0.96||5.56||4.47||<0.001|
This cross-sectional study evaluated for the first time the potential predictors of GE of both solids and liquids in patients with PD. In comparison with an age- and gender-matched control group we have shown that the prevalence of delayed GE is 88% for solids which is in concordance with previous studies9,27,28 and only 38% for liquids in a group of PD patients with mild to moderate disease severity. The observed delay in GE, in particular the lower prevalence of disturbed caloric liquid emptying, is difficult to explain, as an abnormally slow stomach emptying can be considered to result from a variety of factors affecting gastric motor function, e.g. an abnormally high resistance to emptying or a defective mechanical breakdown of solid food, which have not been assessed in this disease so far.13
Gastric emptying of solids was delayed in all untreated PD patients. This might be in concordance with recent neuropathological findings indicating the initial involvement of the dorsal motor nucleus of the vagus nerve in the progressive, neurodegenerative process underlying PD and therefore changes in parasympathetic autonomic supply to the gut could certainly account for the observed delayed GE in formerly untreated and in treated patients.4 In contrast, abnormalities in the enteric nervous system (ENS), which shares embryological, morphological, neurochemical and functional features with the central nervous system,29,30 have also been identified, including both Lewy-body formation and loss of dopaminergic neurones.1,2 However, the role of ENS involvement as a potential novel pathological biomarker has not been established.
For t50 of solids we observed a difference between clinical disease severity groups as assessed by H&Y and UPDRS scores. Despite evaluating all subjects in the ‘defined off’ period and excluding patients with dopaminergic drugs with a half-life of more than 4 h, it remains unclear as to whether these differences are related to the patients’ anti-parkinsonian medication or to different degrees of autonomic or ENS dysfunction. A longitudinal monitoring of gastrointestinal function and its correlation with other biomarkers of PD, such as functional CNS or autonomic nervous system imaging, clinical testing procedures, biochemical and genetic tests, is required to clarify development, extension and progression of gastrointestinal dysfunction in PD and thus to complement the natural history of this disease. In particular, the relationships between UPDRS motor examination subitems rigour and action tremor and GE, which have been observed for the first time in this study, may be clarified in longitudinal observational trials.
Although action tremor and rigour accounted for 46% of the variance of the gastric half-emptying time for solids, as shown by multiple regression analysis, we could not detect any other predictors for GE. Neither gender nor age were associated with the rate of GE in PD patients. While age differences are known to effect solid and liquid GE in health,31 the underlying mechanisms of disturbed gastric motor function in PD might have outweighed these effects, e.g. a putatively impaired accommodation response to a liquid meal in PD and the effect of lower (initial) GE in older subjects.32
In our study approximately 35% of the 80 PD patients reported one or more upper gastrointestinal symptoms, which confirms the known overall prevalence of recurrent upper abdominal discomfort in PD which is higher than in the general population of approximately 25% in the United States and other Western countries.33 However, gastrointestinal symptoms in PD patients, e.g. vomiting and postprandial abdominal fullness, were not associated with a delay in GE in PD patients – unlike in patients with functional dyspepsia or diabetes mellitus, in whom upper GI symptoms predict GE to some extent.34–36 Possible reasons for the observed lack of a relationship in this study could be that gastrointestinal symptoms assessed based on a previous 2-week period under treatment and therefore dopaminergic side effects by dopamine agonists may have induced symptoms like nausea or vomiting, unrelated to a delayed GE.28 In addition, similar to disordered gastrointestinal motility, there are no longitudinal data relating to neurological and gastrointestinal symptoms. As in diabetes, in PD information about the persistence, relapse or the determinants of GI symptom turnover as well as their relationship to the disease process itself could provide important implications for their therapeutic management.4,37
In conclusion, in treated and untreated patients with idiopathic PD solid GE is more likely to be delayed than liquid GE, and the severity of motor handicap, but not any other neurological symptom assessed by UPDRS or upper gastrointestinal symptom or demographic feature, is associated with GE.
Further assessment of gastric motor function in PD is needed to obtain improved understanding of the underlying mechanisms of disturbed gastrointestinal motility and their influence on antiparkinsonian drug distribution/absorption as well as of the mechanisms behind symptom generation.
This study was supported by a local grant (Forschungsförderung an der Medizinischen Fakultät der Ruhr-Universität, FoRUM, Grant to OG and DW). OG is supported by Deutsche Forschungsgemeinschaft (Gö 1358/1–1). The authors thank Monika Kwiatek University Hospital Zurich for her help in manuscript preparation.
- 132005) The physiology of gastric motility and gastric emptying. In: YamadaT, ed. Textbook of Gastroenterology, Vol. 1, pp. 181–206. Lippincott, Philadelphia.. (
- 19Recent developments in Parkinson's disease. In: FahnSMC, CalneDB, GoldsteinM, eds. Recent Developments in Parkinson's Disease, Vol. 2. Florham Park, NJ: Macmillan Health Care Inform, 1987: 153–63., .
- 27Gastric emptying time and gastric motility in patients with untreated Parkinson's disease. Mov Disord 1996; 11: 1041–7., , .
- 30Functional anatomy of the enteric nervous system. In: JohnsonLR, ed. Physiology of the Gastrointestinal Tract, Vol. 1. New York: Raven, 1994: 381–422., , .
- 34Symptoms associated with impaired gastric emptying of solids and liquids in functional dyspepsia. Am J Gastroenterol 2003; 98: 783–8., , , , .Direct Link: