Patients with SSc were recruited from those referred to the Rheumatology Department at the University of Bari Medical School. The diagnosis of SSc fulfilled the American College of Rheumatology (ACR) criteria 423 and the study included 38 consecutive patients (36 females, mean age 50.9 ± 2.0 years, body mass index of 24.6 ± 0.6 Kg/m²). According to ACR classification, 34 patients had limited cutaneous SSc and 4 had diffuse cutaneous SSc. The estimated duration of the disease was calculated from the appearance of the first symptom/sign related to SSc and was 5.8 ± 0.9 years, range 1–21 years. A full range of symptoms was carefully investigated for the cardiovascular system (dyspnea from pericardial effusion, congestive heart failure, palpitations and irregular heart beats due to conduction abnormalities, and congestive heart failure), skin (diffuse pruritus, hyper- or hypopigmentation, and skin tightness and induration), gastrointestinal system (heartburn, dysphagia, hoarseness, constipation alternating with diarrhea, and dyspepsia), respiratory system (dyspnea, chest pain from pulmonary artery hypertension, and cough), musculoskeletal system (arthralgia, myalgia, loss in joint range of motion, and carpal tunnel syndrome symptoms), constitutional (fatigue and weight loss), neurological system (facial pain and hand paresthesias from peripheral entrapment neuropathies, headache, and stroke), genitourinary system (erectile dysfunction and dyspareunia), ears, nose, throat, sicca syndrome, and endocrine system (hypothyroidism). Apart from skin fibrosis, further clinical features of the patients investigated included: Raynaud's phenomenon (100%), arterial hypertension (26%), pulmonary interstitial fibrosis (3 out 4 patients and 26% in diffuse cutaneous SSc and limited cutaneous SSc, respectively), hypertension (2 out 4 patients and 37% in diffuse cutaneous SSc and limited cutaneous SSc, respectively), and esophageal dysmotility (95%). All patients underwent gastric endoscopy and were on antireflux therapy with daily doses of proton pump inhibitors.

A group of 60 matched healthy volunteers without symptoms, medical treatments or a prior history of disease (33 females, mean age 47 ± 2 years, body mass index 23.9 ± 0.5 Kg/m²) served as healthy controls and 68 matched subjects with no other disease except for functional dyspepsia (45 females, mean age 49 ± 2 years, body mass index 24.9 ± 0.5 Kg/m²) served as dyspeptic controls. Exclusion criteria for all subjects enrolled in the study were the presence of gallstones, acute or chronic pancreatitis, diabetes, pregnancy, peptic ulcer disease, history of esophageal and abdominal surgery and any other diseases or pharmacological treatment (including prokinetics and calcium channel blockers) known to affect gastrointestinal motility. Exclusion criteria also included the presence of contraindications for the entire series of cardiovascular tests, i.e. arrhythmia, recent myocardial infarction, drug therapy that could interfere with cardiovascular activity, severe metabolic disturbance and unwillingness to cooperate. In addition, subjects were excluded if heavy smokers (i.e. more than 10 cigarettes/day), alcohol (i.e. more than 20 g ethanol/day in both male and females, respectively) or drugs abusers. Scleroderma patients who had received treatment other than small daily doses of steroids, ACE-inhibitors for arterial hypertension and/or d-penicillanine were excluded. All patients were evaluated within the yearly scheduled routine work-up which included blood tests and abdominal ultrasonography. Approval from the institutional review board and written informed consent from the participating subjects were obtained.

After completing the full work-up at the Rheumatology Section, SSc patients were asked to join the study on gastrointestinal symptoms, motility, and autonomic nervous system function at the Section of Internal Medicine. All enrolled subjects underwent complete clinical examination, an interview by questionnaires for the evaluation of dyspepsia, upper gastrointestinal perception, and bowel habits, functional ultrasonography for the study of gallbladder and gastric motility, H₂-breath test for orocecal transit time (OCTT). SSc patients also underwent cardiovascular tests for autonomic nervous system function (see below). Tests were started at 08:00 AM. Although the maximal length of the work-up was set at 5 hours, the true length was invariably less than 3 hours.

The presence of dyspepsia was assessed and quantified by a validated semi-quantitative scoring system measuring a subset of four symptoms (epigastric pain, burning, belching/burping, postprandial fullness), resulting in a maximal score equal to 48 24. In this series, the upper normal limit estimated from the mean + 2SDs of healthy controls was equal to 8 2526.

On the day of the motility studies, appetite, satiety, nausea, epigastric fullness and epigastric pain (or discomfort) were monitored by the visual analogue scale (VAS) 27. The VAS consisted of a 100 mm-horizontal line indicating the current intensity of the perception between two extremes (extreme left = "no symptom at all"; right end = "maximum feeling of..."). This line was presented to the subjects who were asked to draw a vertical line through the horizontal line along its length. The distance in mm from the extreme left of the line to the vertical line is a quantitative measure of gastrointestinal perception 28. Scores at baseline (i.e. time 0) was the mean of 3 measurements over 3 days in the fasting state at the same time in the morning. Afterwards, VAS was scored at 15, 30, 45, 60, 90 and 120 minutes postprandially. The VAS has been previously demonstrated as a reliable method to assess subjective perception in response to experimental stimuli in the upper gastrointestinal tract 12272829.

The weekly frequency of bowel movements was calculated from a daily diary over a time span of one month. The "Bristol" stool form scale was used to assess the quality of stools based on a semiquantitative scale, as a marker of colonic transit. The scale is deemed to be a reliable method to estimate colonic transit time 3031.

Fasting and postprandial course of gastric antral area and gallbladder volume were assessed by functional ultrasonography using a 3.5 MHz convex probe attached to an Ansaldo-Hitachi equipment, as previously described also by our group 283233. The test meal was a 200 mL solution with 13 g (39%) fat, 10 g (13%) protein and 35 g (48%) carbohydrates for a total of 300 kcal, 1270 kJ, 365 mOsm/l (Nutridrink^®, Nutricia, Milano, Italy). The drink was taken at room temperature over one minute in the presence of the examinator. Indices of gastric emptying were: fasting antral area (mean of 3 measurements at -15, -5 and 0 minute before test meal, expressed in cm²); maximal antral area at 2 minute post-meal; minimal postprandial antral area during the 2-hour; half-emptying time. Gastric emptying curves were obtained by plotting antral areas vs. time. In order to obviate inter-individual variability of antrum size 32, postprandial areas were normalized to maximal area after subtracting basal areas: i.e. 100 × (A_t-a)/(A₂-a), where A_t = postprandial area at any given time; a = basal area; A₂ = area at 2 minute postprandially 34. Half-emptying time was the time at which 50% decrease of antral area occurred (T_1/2, minute), and calculated by linear regression analysis from the linear descending part of the antral emptying curve. This index correlates closely with the scintigraphic T_1/2 28. Gastric emptying was considered abnormal if T_1/2 was greater than 35 minutes (mean + 2 SD in healthy controls) 2526.

Gallbladder emptying was performed simultaneously to gastric motility by the same operator. Fasting and post-meal gallbladder volumes were measured by sagittal and transverse scans of the gallbladder at its largest dimension at 5–15 minute intervals during 2 hours. Gallbladder volume was measured from frozen sonograms assuming an ellipsoid shape of the organ – a reliable method when gallbladder shape is not highly irregular 35, which was the case in this study. In healthy subjects the liquid meal employed in this test (see above) induces more than 50% emptying of fasting gallbladder volume, and has been validated by our group on several occasions 2526333637. Indices of gallbladder motility were as follows: fasting volume (mean of 3 measurements at -15, -5 and 0 minute before test meal, expressed in mL); residual volume (minimal volume measured postprandially, expressed in mL and as percent of fasting volume); T_1/2 (time to achieve 50% decrease of fasting volume), calculated by linear regression analysis from the linear descending part of the emptying curve 36. Gallbladder emptying was considered abnormal if half-emptying time was greater than 35 minutes (mean + 2 SD in healthy controls) 2526.

OCTT was measured by the hydrogen (H₂) breath test 38. During the 5 days before the test, subjects were not allowed to take antibiotics, probiotics, prokinetics or other drugs known to affect gastrointestinal motility. To avoid prolonged intestinal H₂-production secondary to the presence of non-absorbable or slowly fermentable material in the colonic lumen, a special diet was given starting the day before the test and consisting of proteins and fat (meat, fish, eggs, olive oil) 252639. The meal was followed by a 12-hour fasting period. Between 08:00 and 09:00 AM the test was started. Smoking and physical exercise were prohibited one hour before and during the test 4041. Two breath samples were taken in the fasting state and afterwards the subject ingested 10 g of lactulose (Duphalac Dry^®, Solvay Pharma, Torino, Italy) added directly as a powder to the liquid test meal used for simultaneous study of gallbladder and gastric emptying. End-respiratory breath samples were analyzed every 10 minutes for five hours after lactulose ingestion, using a pre-calibrated, portable hydrogen sensitive electrochemical devices (EC60-Gastrolyzer, Bedfont, USA and Lactofan H₂ Analyzer, Fisher Analysen Instr. Leipzig, Germany, kindly donated by Italchimici SpA, Pomezia, Rome, Italy). The concentration of hydrogen was measured in parts per million (ppm), and the accuracy of the detector was ± 2 ppm. A rise of 10 ppm above baseline on two consecutive measurements was considered as OCTT and expressed in minute 38. In all subjects a hydrogen peak was evident, thus excluding the presence of non-H₂ producers. OCTT was considered delayed if greater than 120 minutes (i.e. the mean+2SD obtained by a large healthy control group) 2526. The presence of small intestinal bacterial overgrowth was excluded by H₂ breath test as previously validated 42 and according to published criteria 43. The lactulose breath test was considered normal if there was no rise in H₂ concentration during the first 90 minutes after lactulose ingestion, with a definitive rise never greater than 20 ppm during 180 minutes of measurement 44.

The integrity of the autonomic nervous system was assessed by both the Sweat-Spot-Test (SST) 384546 and cardiorespiratory reflex tests 47, as previously described by our group 26. The SST was used to investigate the presence of cholinergic sympathetic fibers alterations by analyzing sweat abnormalities on the skin of the dorsum of the foot, which was coated with iodine and a fine emulsion of starch in arachis oil. Sweat was stimulated either by an intra-dermal injection of 0.1 mL acetylcholine (Ach.SST), or a standardized thermal stimulus which includes pre-ganglionic fibers (Thermal SST). A colorimetric reaction between starch and iodine was triggered by the sweat from stimulated glands, so that each pore was seen as a little black dot after 2–5 minutes. A digital photo was obtained and transferred to a magnifying software to measure the number and distribution of dots appearing in a standard squared grid of 529 mm² divided into 64 squared subareas. A normal SST was expressed by a score ≥ 12 dots/subarea and/or < 8% of abnormal subareas (each square of the grid having less than 6 dots), according to Ryder 45 and to our group 2646. Only patients with both indices (SST score and % of abnormal subareas) outside normal limits were considered to have a positive test. A portable device (Cardionomic^®, Lifescan, Italy) connected to skin electrodes was employed to measure the "beat-to-beat" modifications of the R-R interval. Lying-to-standing and standing-to-lying tests were used to assess sympathetic involvement. Valsalva maneuver, deep-breathing, cough test, and postural hypotension test were used for parasympathetic involvement. The results from each test were scored as normal, borderline or abnormal, according to age-controlled values. The overall results for autonomic neuropathy were therefore normal (all tests normal), early involvement (one abnormal test or two borderline abnormal), definitive involvement (two or more abnormal tests) 2648.

Calculations were performed with the NCSS 2007 statistical software (Hintze J, 2006, Kaysville, UT, USA), and data expressed as the mean ± standard error (SE). Continuous variables were analyzed using ANOVA followed by Fisher's LSD test. Differences between two subgroups were evaluated by Mann-Whitney U-test or by Student's t test, where appropriate. Linear regression analysis was performed by the method of least square. The chi-square test was used to compare categorical variables and proportions. A two-tailed probability (P) value of less than 0.05 was considered statistically significant 4950.

The score for dyspepsia was abnormal in 61% of SSc patients, with a mean score greater than healthy controls, but lower than dyspeptic controls (Table 1). This was also seen when dyspeptic controls were compared with scleroderma patients with abnormal dyspeptic score (n = 23, mean score 17.3 ± 1.5, P = 0.0009). Estimated disease duration was significantly (P = 0.03) longer in SSc patients with abnormal dyspepsia score (7.3 ± 1.4 years) than in those without dyspepsia (3.4 ± 0.6 years).

The analysis of dyspeptic symptoms by VAS throughout the motility study showed that the feeling of appetite and satiety were similar in SSc patients, healthy and dyspeptic controls. As expected, the perception of satiety and appetite (either in fasting and postprandial period) showed a strong negative correlation in both patients and controls (r = -0.80; P = 0.000001).

Dyspeptic controls had more nausea than both SSc patients and healthy controls (Table 1).

Of note, perception of epigastric fullness in SSc patients was constantly higher than in healthy controls, although dyspeptic controls had the highest score for this symptom (Table 1 and Figure 1). Finally, dyspeptic controls showed higher perception of epigastric pain than both SSc patients and healthy controls (Table 1).

Despite the low number of patients with diffuse SSc enrolled in the present study, no difference was evident between patients with limited or diffuse scleroderma for either dyspepsia score (12.8 ± 1.5 and 11.0 ± 2.3, respectively, P = NS) or dyspeptic symptoms by VAS (data not shown).

Twenty-nine out of 38 patients and all of controls gave their informed consent to undergo tests for autonomic neuropathy. In terms of clinical features, this subgroup was fully representative of the whole SSc group of patients (n = 25 with limited and n = 4 with diffuse disease). Overall, tests for autonomic neuropathy were abnormal in 20 (69%) of the patients but in none of the controls. Abnormal parasympathetic function, sympathetic function or both were found in 79%, 55% and 41% of cases, respectively. According to disease classification, tests for autonomic neuropathy were abnormal in 15 out of 25 patients (60%) with limited disease and in 1 out of 4 subjects (25%) with diffuse scleroderma. No significant differences could be detected in patients with respectively normal or abnormal tests for dyspeptic scores (13.0 ± 3.3 and 11.6 ± 1.7,), bowel habits (Bristol score 3.7 ± 0.6 and 3.4 ± 0.3), fasting gallbladder volume (22.5 ± 2.3 and 25.2 ± 2.1 mL), residual gallbladder volume (6.4 ± 1.8 and 5.7 ± 0.7 mL), fasting postprandial antral areas (2.8 ± 0.1 and 2.9 ± 0.2 cm²), postprandial antral areas (3.4 ± 2.1 and 5.1 ± 1.4%), antral half-emptying time (36.7 ± 4.4 and 34.3 ± 2.4 min), and OCTT (190.0 ± 16.0 and 158.3 ± 11.3 minutes).