A Stable Isotope Breath Test With a Standard Meal for Abnormal Gastric Emptying of Solids in the Clinic and in Research
2008; Elsevier BV; Volume: 6; Issue: 6 Linguagem: Inglês
10.1016/j.cgh.2008.01.009
ISSN1542-7714
AutoresLawrence A. Szarka, Michael Camilleri, Adrian Vella, Duane D. Burton, Kari Baxter, Julie Simonson, Alan R. Zinsmeister,
Tópico(s)Dysphagia Assessment and Management
ResumoBackground & Aims: The aim of this study was to validate a [13C]–Spirulina platensis gastric emptying (GE) breath test (GEBT) with a standardized meal. Methods: Thirty-eight healthy volunteers and 129 patients with clinically suspected delayed GE underwent measurements at 45, 90, 120, 150, 180, and 240 minutes after a 238 kcal meal labeled test with 100 mg [13C]–S platensis and 0.5 mCi 99mTc. We established normal ranges for scintigraphy with this test meal, intraindividual and interindividual coefficients of variation (COVs), and the ability of the [13C] GEBT breath percent dose excreted *1000 values to predict scintigraphic half-life and to categorize GE as delayed, normal, or accelerated. Results: In health, the 10th and 90th percentiles of half-life for scintigraphic GE with this meal were 52 and 86 minutes; intraindividual COVs for scintigraphy and the GEBT were, respectively, 31% and 27% at 45 minutes, 17% and 21% at 90 minutes, 13% and 16% at 120 minutes, 10% and 13% at 150 minutes, and 8% and 12% at 180 minutes. Interindividual COVs at each time for the [13C] GEBT and scintigraphy were typically ∼1%–4% lower than intraindividual COVs. Individual breath samples at 45, 150, and 180 minutes predicted GE category; at 80% specificity, 45- and 180-minute samples combined were 93% sensitive to identify accelerated GE, and 150- and 180-minute combined were 89% sensitive for delayed GE. Conclusions: [13C]–S platensis GEBT is as reproducible as scintigraphy; imprecision with both tests reflects physiologic variation. With 4 breath samples, this method with an off-the-shelf meal is valid to assess GE in clinic and in research. Background & Aims: The aim of this study was to validate a [13C]–Spirulina platensis gastric emptying (GE) breath test (GEBT) with a standardized meal. Methods: Thirty-eight healthy volunteers and 129 patients with clinically suspected delayed GE underwent measurements at 45, 90, 120, 150, 180, and 240 minutes after a 238 kcal meal labeled test with 100 mg [13C]–S platensis and 0.5 mCi 99mTc. We established normal ranges for scintigraphy with this test meal, intraindividual and interindividual coefficients of variation (COVs), and the ability of the [13C] GEBT breath percent dose excreted *1000 values to predict scintigraphic half-life and to categorize GE as delayed, normal, or accelerated. Results: In health, the 10th and 90th percentiles of half-life for scintigraphic GE with this meal were 52 and 86 minutes; intraindividual COVs for scintigraphy and the GEBT were, respectively, 31% and 27% at 45 minutes, 17% and 21% at 90 minutes, 13% and 16% at 120 minutes, 10% and 13% at 150 minutes, and 8% and 12% at 180 minutes. Interindividual COVs at each time for the [13C] GEBT and scintigraphy were typically ∼1%–4% lower than intraindividual COVs. Individual breath samples at 45, 150, and 180 minutes predicted GE category; at 80% specificity, 45- and 180-minute samples combined were 93% sensitive to identify accelerated GE, and 150- and 180-minute combined were 89% sensitive for delayed GE. Conclusions: [13C]–S platensis GEBT is as reproducible as scintigraphy; imprecision with both tests reflects physiologic variation. With 4 breath samples, this method with an off-the-shelf meal is valid to assess GE in clinic and in research. The measurement of gastric emptying (GE) by stable isotope breath tests (GEBTs) has practical and safety advantages compared with current scintigraphic methods.1Ghoos Y.F. Maes B.D. Geypens B.J. et al.Measurement of gastric emptying rate of solids by means of a carbon-labeled octanoic acid breath test.Gastroenterology. 1993; 104: 1640-1647Abstract PubMed Google Scholar Unlike scintigraphy, which requires elaborate detection equipment and the patient to be located in the same setting, GEBT can be performed just about anywhere, including any office or bedside, because the collected breath samples are stable, and the samples can sent to a remote site for analysis. GEBT is safer than scintigraphy because it involves no radiation exposure, which is advantageous if repetitive assessments of GE are needed for research or clinical purposes, or if assessment of GE is needed in pregnant or breast-feeding women and in children. In previous studies conducted in our laboratory,2Choi M.-G. Camilleri M. Burton D.D. et al.13C-octanoic acid breath test for gastric emptying of solids: accuracy, reproducibility and comparison with scintigraphy.Gastroenterology. 1997; 112: 1155-1162Abstract Full Text PDF PubMed Scopus (179) Google Scholar, 3Choi M.-G. Camilleri M. Burton D.D. et al.Reproducibility and simplification of 13C-octanoic acid breath test for gastric emptying of solids.Am J Gastroenterol. 1998; 93: 92-98Crossref PubMed Scopus (119) Google Scholar, 4Lee J.S. Camilleri M. Zinsmeister A.R. et al.A valid, accurate, office based non-radioactive test for gastric emptying of solids.Gut. 2000; 46: 768-773Crossref PubMed Scopus (72) Google Scholar, 5Lee J.-S. Camilleri M. Zinsmeister A.R. et al.Toward office-based measurement of gastric emptying in symptomatic diabetics using [13C]octanoic acid breath test.Am J Gastroenterol. 2000; 95: 2751-2761Crossref PubMed Google Scholar we demonstrated that as compared with simultaneous scintigraphy, the 13-carbon ([13C])-octanoate and [13C]–Spirulina platensis GEBT provided an acceptable assessment of the GE of solids in humans, with acceptable coefficient of variation (COV) comparable to scintigraphy. [13C]–S platensis GEBT was able to identify accelerated or delayed emptying induced pharmacologically with placebo, erythromycin, or atropine. Across the range of GE, the mean difference in half-life between the 2 methods was 0.15 minutes with standard deviation of 35.5 minutes.6Viramontes B.E. Kim D.-Y. Camilleri M. et al.Validation of a stable isotope gastric emptying test for normal, accelerated or delayed gastric emptying.Neurogastroenterol Motil. 2001; 13: 567-574Crossref PubMed Scopus (68) Google Scholar The meals used in the prior studies were not completely standardized (eg, egg size and weight might differ) and were not shelf-stable. To facilitate safe, point-of-care assessment of GE, a standardized test meal consisting entirely of shelf-stable components including 100 mg [13C]–S platensis has been developed.7[C13]-Spirulina platensis gastric emptying breath test (GEBT): investigator brochure. Advanced Breath Diagnostics, LLC, Brentwood, TN2005Google Scholar In the current prospective, validation study comparing [13C]–S platensis GEBT by using a standardized, shelf-stable meal with simultaneous scintigraphic GE, our aims were (1) to establish normal ranges for scintigraphy with this test meal, (2) to appraise the performance characteristics (intraindividual and interindividual COVs of both scintigraphy and [13C]–S platensis GEBT) in healthy volunteers, and (3) to assess the ability of the [13C]–S platensis GEBT breath (percent dose excreted *1000 [kPCD]) values to predict scintigraphic half-life (t1/2) and to categorize GE as delayed, normal, or accelerated in patients with symptoms suggestive of abnormal GE. This prospective, open-label comparison validation study was conducted at the Mayo Clinic's Clinical Research Unit. All studies were approved by the Mayo Clinic Institutional Review Board. In the first phase, we performed a calibration study in 38 healthy volunteers to estimate the reference range for scintigraphic results by using the standardized, shelf-stable test meal and to estimate the total variability (imprecision) of scintigraphy and [13C]–S platensis GEBT measurements on healthy volunteers. Twenty-eight participants underwent studies on 2 occasions to study intraindividual variation. In the second phase of this study, we validated the [13C]–S platensis GEBT for use in the diagnosis of delayed GE by studying 124 participants who were referred for scintigraphy for clinically suspected abnormal GE and in 5 healthy subjects who received atropine (0.01 mg/kg IV bolus during a period of 10 minutes followed by 0.01 mg/kg infusion during a period of 50 minutes) to pharmacologically delay GE. In the first phase, we recruited by public advertisement 38 normal participants, men and women between 18–75 years old. Women of childbearing potential were required to have negative pregnancy urine test within 48 hours of the dual-label GE test. Participants were excluded if they had any history, physical examination, or laboratory finding to suggest systemic diseases, history of abdominal surgery except appendectomy; clinically significant neurologic or psychiatric disorders, use of narcotics or anticholinergic agents within 2 days of the study, or receipt of an investigational drug within 4 weeks before the study. In the second phase of the study, 124 participants were recruited from patients referred for GE by scintigraphy on the basis of clinical assessment by physicians at the Mayo Gastroenterology Motility Clinic. Participants were men and women, 18–75 years old, and had similar inclusion criteria, except concomitant general diseases or suspicion of delayed GE were not exclusion criteria, and in addition, there could be no history or suspicion of malabsorption caused by mucosal disease, pancreatic disease, or liver dysfunction. The 5 normal participants, who received atropine in the second phase of the study, were recruited by public advertisement and had the same inclusion/exclusion criteria as the normal participants used in the first phase of the study. For each potential participant, a screening visit was conducted in which consent was obtained, and a physical exam was performed. After an overnight fast (minimum 8 hours), the participants returned to the study center at approximately 7:00 am, at which time the dual-label GE test was started. The patient consumed the test meal containing [13C]–Spirulina and 99mTc sulfur colloid. Scintigraphic images were obtained on completion of the meal and at 45, 90, 120, 150, 180, and 240 minutes after the meal. Breath samples were collected at baseline before the test meal was started and simultaneously with scintigraphic image acquisitions after the ingestion of the test meal. The test meal consisted of 1 mCi 99mTc-sulfur colloid, 100 mg [13C]–S platensis, 27 g freeze-dried egg mix, 6 saltine crackers, and 180 mL of water. The caloric content of the meal is 238 kcal, and the meal has a balanced composition of 16.9 g carbohydrates, 14.4 g protein, and 11.2 g fat. The nature and size of the meal were selected to ensure stability at room temperature, palatability, and calorie content that would be consumed entirely, even by patients with suspected gastroparesis and upper abdominal symptoms. S platensis is a protein-rich, blue-green algae eaten as a food source in many parts of the world and is sold as a dietary supplement in the United States.8Ciferri O. Spirulina, the edible microorganism.Microbiol Rev. 1983; 47: 551-578PubMed Google Scholar, 9FDA talk paper. "Spirulina". June 23, 1981.Google Scholar, 10Ciferri O. Tiboni O. The biochemistry and industrial potential of Spirulina.Ann Rev Microbiol. 1985; 39: 503-526Crossref PubMed Scopus (159) Google Scholar It contains 50%–60% protein, 30% starch, and 10% lipid.11Dillon J.C. Phuc A.P. Dubacq J.P. Nutritional value of the alga Spirulina.World Rev Nutr Dietetics. 1995; 77: 32-46PubMed Google Scholar The natural level of 13C in S platensis and in all living things is about 1%.12Ricci E. Determination of carbon-12, carbon-13 isotopic abundances and nitrogen-carbon ratios in biological substances by proton-reaction analysis.Analyt Chem. 1971; 43: 1866-1871Crossref PubMed Scopus (22) Google Scholar The S platensis used in this study was grown in a closed hydroponics chamber charged with pure 13C-source, raising the level of 13C in the resultant cells to 99%.7[C13]-Spirulina platensis gastric emptying breath test (GEBT): investigator brochure. Advanced Breath Diagnostics, LLC, Brentwood, TN2005Google Scholar Because the contents of the algal cells are not freely diffusible, incorporation of 13C-labeled S platensis into the egg mix provides a way to assess the emptying of the solid phase of the meal. 13C can only be released from the algal cells after the egg mix is emptied from the stomach, the cells are digested, and the 13C-labeled substrates (algal protein, fat, and carbohydrate) are absorbed and metabolized. In this way, [13C]-S platensis gives rise to respiratory CO2 that is enriched in 13C. Breath samples were taken at baseline before the meal and followed the same time schedule as the scintigraphic technique. End-tidal breath samples were collected while the participant's abdomen was being imaged by the gamma camera. Breath samples were stored in duplicate in glass screwcap Exetainer tubes (Labco Limited, High Wycombe, UK) by using a straw to blow into the bottom of the tube to displace contained air. After recapping the tubes, the 13CO2 breath content was determined in a centralized laboratory (AB Diagnostics, Brentwood, TN) by isotope ratio mass spectrometry. The 13C enrichment was expressed as the delta per mL difference between the 13CO2/12CO2 ratio of the sample and the standard. To calculate the quantity of 13C appearing in breath per unit time, delta over baseline (DOB) was used where 0.0112372 is the isotopic abundance of the limestone standard, Pee Dee belemnite, and CO2 production was corrected for age, sex, height, and weight by using the algorithms of Schofield et al as described by Klein.13Klein P.D. Clinical applications of 13CO2 measurements.Federation Proceedings. 1982; 41: 2698-2701PubMed Google Scholar The currently preferred GEBT metric is the percent dose (PCD) excreted at time t after consumption of the test meal.14Schoeller D.A. Schneider J.F. Solomons N.W. et al.Clinical diagnosis with the stable isotope 13C in CO2 breath tests: methodology and fundamental considerations.J Lab Clin Med. 1977; 90: 412-421PubMed Google Scholar To provide a more convenient scale, we multiplied PCD by 1000 to produce kPCD at any time, t.kPCDt=[DOB∗CO2PR∗Rs∗1310∗dose]∗1000 where DOB is the measured difference in the ratio [13CO2/12CO2] between a post-meal breath specimen at any time (t minutes) and the baseline breath specimen. CO2 production rate (CO2PR) (mmol CO2/min) was calculated by using the equations of Schofield,15Schofield W.N. Predicting basal metabolic rate, new standards and review of previous work.Hum Nutr Clin Nutr. 1985; 39: 541-549Google Scholar which incorporate the patient's age, gender, height, and weight. Rs is the ratio [13CO2/12CO2] in the reference standard (Pee Dee belemnite) for these measurements, Rs is 0.0112372, 13 is the atomic weight of carbon-13, 10 is a constant factor for converting units, and dose is the weight (mg) of carbon-13 in the dose of [13C]-S platensis administered to the patient in the test meal. Because [13C]–S platensis is approximately 43% carbon-13, a dose of 100 mg [13C]–S platensis corresponds to approximately 43 mg of carbon-13. A region of interest (ROI) was drawn around the stomach on the anterior and posterior images for each time frame. Data were corrected for decay of 99mTc. To correct for depth or tissue attenuation, the counts of each anterior and posterior ROI were multiplied together, and the square root of the product was taken to obtain the geometric mean. The scintigraphic GE metric, Propt, is the proportion of tracer emptied from the stomach at time, t. With a power exponential model, these data were also used to calculate the GE t1/2 after estimating the constants κ and β in the power exponential model, Propt = exp(−κtβ). Data from the first phase of this study were summarized separately for initial and repeat studies over all subjects. Repeat studies in 28 of the 38 subjects were used to compute intraindividual COVs at each scan/breath sample time point (100*SD of the deltas/overall grand mean). The individual proportions (Propt) and the calculated t1/2 values obtained from the scintigraphic data of the first study in all 38 normal volunteers were summarized. In particular, the 10th and 90th percentiles for these data were used to define cutoffs indicating delayed GE (eg, t1/2 >90th percentile), accelerated GE (eg, t1/2 <10th percentile), with the remainder considered as normal GE (eg, between 10th and 90th percentiles). We selected 10th and 90th percentiles rather than the customary normal range of 5th–95th percentiles because this is more likely to be representative of a normal range based on n = 38. These cutoffs were then used to categorize subjects into the corresponding 3 groups by using the scintigraphically derived t1/2 values for subjects from the second phase of the study. The joint association between scintigraphic GE proportions at 45, 90, 120, 150, and 180 minutes and the breath kPCD values at the same time points was assessed with a canonical correlation analysis that essentially tests simultaneously the 5 × 5 correlation matrix between the 2 sets (proportions of GE and kPCD values at each of 5 time points) of variables. This procedure generates up to 5 canonical variates, which are linear combinations of the scintigraphic and BT variables; the first canonical variate for each set is chosen to reflect the maximal (linear) correlation among all possible linear combinations of each set of initial variables. The first canonical variate for each of the 2 sets was plotted to illustrate the correlation between the 2 sets of variables. In view of the results available from the canonical correlation analysis, multiple linear regression models to predict the individual scintigraphic GE proportions at each time point by using the set of 5 kPCD values were developed. These models incorporated gender and body mass index (BMI) as covariates. From the predicted GE proportions, a BT estimate of the corresponding t1/2 value could be computed by using a simple linear interpolation algorithm (ie, linearly interpolate between the predicted GE proportions around 0.5, that is for emptying of 50% of the meal, to correspond with the t1/2). The concordance between the scintigraphic t1/2 values and the BT-estimated t1/2 values was then estimated,16Carrasco J.L. Jover L. Estimating the generalized concordance correlation coefficient through variance components.Biometrics. 2003; 59: 849-858Crossref PubMed Scopus (181) Google Scholar and a Bland-Altman plot was generated to determine whether this would be a useful method to estimate GE t1/2 values for use in clinical practice or research. The ability of the breath kPCD values to predict scintigraphic t1/2 values was also assessed by using a logistic regression model (with the 3-group category, delayed, accelerated, and normal, as the dependent variable) and the breath kPCD values as predictors. This approach provides a diagnostic method useful in clinical practice. A backward elimination approach was used to identify the best subset of kPCD values to predict the t1/2 category of each subject. The predicted probabilities (for delayed versus normal and for accelerated versus normal) for each subject were then used as the marker values to generate receiver operating characteristic (ROC) curves. The ROC curves and corresponding area under curve (AUC) statistics were obtained for the best subset model as well as for some single BT models and some models based on 2 BT time points. The SAS software package (SAS Institute Inc, Cary, NC) was used for all statistical analyses.17SAS Institute IncSAS/STAT users guide, version 6.in: 4th ed. SAS Institute, Cary, NC1989: 1135-1194Google Scholar The study was designed to provide adequate power to test the sensitivity and specificity of the breath kPCD values to identify delayed GE separately at each of several a priori chosen time points in a representative sample of patients presenting with upper gastrointestinal symptoms that might be attributable to abnormal GE. To test whether a specific cutoff in breath kPCD values at any particular time point had a sensitivity of 0.95 (null hypothesis) versus the alternative hypothesis that the sensitivity was less than 0.85, a sample size of 45 subjects with delayed GE (by scintigraphic methods [gold standard]) would be needed for 80% power (1-sided alpha level of .025). Historical data at the Mayo Clinic Rochester Motility Clinic indicate that approximately 35% of adults examined by gastric scintigraphy have delayed GE. This implied 45/0.35 (roughly 128) total subjects would be needed. Similarly, if the subjects were characterized on the basis of a composite of the scintigraphic proportions emptied (eg, a GE t1/2 value) as delayed versus normal versus accelerated, then a sample of 45 with delayed t1/2 values would be needed to test the sensitivity of a composite BT score cutoff (eg, a sensitivity of 0.95 versus less than 0.85 for delayed versus normal t1/2 values). Alternatively, to develop a logistic regression model with 5 predictor variables (ie, the kPCD values at 5 time points) to discriminate between delayed and normal GE would require about 50 subjects in the smaller of the 2 groups (delayed versus normal GE). The demographics of the participants in both the calibration phase (health) and the validation phase (patients) of the study and the referral diagnosis for GE assessment are shown in Table 1, Table 2.Table 1Subjects in Calibration Study (Phase 1) to Establish Reference Range in HealthMenWomenNumber1127Median age (y and IQR)21 (19–21.9)37 (20.8–45.5)Median BMI (kg/m2 and IQR)26 (24.1–27.2)23 (21.0–26.9)IQR, interquartile range. Open table in a new tab Table 2Subjects in Prospective Validation Study (Phase 2: 124 Patients, 5 Healthy Subjects): Demographics and DiagnosisMenWomenNumber3693Median age (y and IQR)49.7 (34.5–61.0)49.9 (37.3–61.3)Median BMI (kg/m2 and IQR)28.3 (24.6–30.7)26.8 (22.2–29.8)Functional dyspepsia1142Gastroparesis (nondiabetic)44Diabetes56Connective tissue disorders02Other motility disorders44Gastroesophageal reflux disease65Status post fundoplication32Idiopathic nausea220Other13Healthy (received atropine)05IQR, interquartile range. Open table in a new tab IQR, interquartile range. IQR, interquartile range. Table 3 shows the data from 38 healthy volunteers (27 women, 11 men), and the 10th and 90th percentiles for the GE t1/2 for this meal as measured by scintigraphy in healthy volunteers were 52 and 86 minutes, respectively. Replicate data were acquired on 2 occasions in 22 female and 6 male healthy volunteers. The intraindividual COV (SD of deltas/overall mean) is 26% for the scintigraphic GE t1/2 values, with a 95% confidence interval for the mean delta of −7 minutes to 8 minutes. The (linear) concordance coefficient for the pairs of t1/2 values was 0.50. Intraindividual COVs for scintigraphy and the [13C]-S platensis GEBT were, respectively, 31% and 27% at 45 minutes, 17% and 21% at 90 minutes, 13% and 16% at 120 minutes, 10% and 13% at 150 minutes, and 8% and 12% at 180 minutes. Interindividual COVs for scintigraphy and the GEBT were, respectively, 26.7.% and 21.3.% at 45 minutes, 16.2% and 18.9% at 90 minutes, 12.1% and 15.7% at 120 minutes, 9.3% and 13.9% at 150 minutes, and 6.0% and 11.8% at 180 minutes.Table 3GE Characteristics at Different Time Points in 38 Healthy Volunteers in First Phase of StudyMeanSD10th percentile90th percentileAge (y)31.7514.7018.4654.56BMI (kg/m2)24.534.15419.9431.99T½ (min)67.7915.1152.4385.75GE 45 min0.3360.0900.2050.441GE 90 min0.6480.1050.5280.778GE 120 min0.8000.0970.6950.934GE 150 min0.8900.0830.7800.993GE 180 min0.9480.0570.8701.000GE 240 min0.9910.0200.9671.000NOTE. Data show proportion emptied (GE) at specified times.SD, standard deviation. Open table in a new tab NOTE. Data show proportion emptied (GE) at specified times. SD, standard deviation. Figure 1A shows a summary of the GE results by t1/2 group in the 129 subjects with clinically suspected delayed GE. The data for each point are provided in detail in Supplemental Table 1. Note that 55 of 129 (42.6%) had delayed GE and 14 of 129 (11%) patients had accelerated GE. Figure 1B shows examples of plots of observed GE (proportions remaining in the stomach) and predicted proportions (based on equations in the text) by using the breath kPCD values shown (along with BMI and gender). The BT-estimated t1/2 values are then computed by linear interpolation of the predicted proportions just above and below 0.5.Supplementary Table 1Summary of GE Results by t1/2 Group in 129 Subjects With Clinically Suspected Delayed GE (n = number)GroupProportion emptied at time pointMeanSDNormal t½, 52–86 min (n = 60)GE450.310.06GE900.600.09GE1200.750.08GE1500.860.07GE1800.940.05Accelerated t½, 86 min (n = 55)GE450.180.08GE900.330.13GE1200.450.16GE1500.570.19GE1800.680.21 Open table in a new tab Table 4 demonstrates the excellent correlations at individual time points for the GE parameters by scintigraphy and the corresponding time kPCD values by BT. These data essentially establish the strong association between scintigraphic GE proportions and [13C]–S platensis GEBT breath kPCD values and support the rationale for developing a simple analysis of the BT for its application in practice. The canonical correlation analysis indicated 3 significant pairs of canonical variates, and the first pair is plotted in Figure 2. These canonical variates are (weighted) linear combinations of the GE proportions and breath kPCD values, respectively. They are thus related to other specific functions of the GE proportions. Specifically, the first canonical variate for the GE proportions (Y-axis in Figure 2) has a strong (negative) relationship with the scintigraphically based t1/2 values (Spearman correlation of −0.82, P < .001). The first canonical variate of the kPCD values (X-axis in Figure 2) was also strongly related to these t1/2 values (Spearman correlation of −0.73, P < .001). Because slower GE corresponds to larger t1/2 values, the correlation of these canonical variates with t1/2 values is negative, reflecting the propensity of the slower GE subjects in the lower left quadrant and the faster GE subjects in the upper right quadrant.Table 4Correlations Between Proportion Emptied From the Stomach and kPCD by [13C]–S Platensis GEBT at a Priori Chosen Time Points in all ParticipantsPearson correlation coefficients, n = 129BT45BT90BT120BT150BT180GE450.628360.542180.482580.367260.25435GE900.741210.814130.781990.673230.52818GE1200.737000.818070.833100.766500.64339GE1500.716350.791030.836050.824500.73545GE1800.682970.725190.784330.801730.77191NOTE. All correlations are P < .001 except for correlation between GE 45 and BT 180, which was .004. Note that, in general, the largest correlation coefficients for GE scintigraphy and BT occur at the times when sampling occurred (eg, 45, 90, 120 minutes). However, later in the study (150 and 180 minutes), the correlations are not necessarily matched for time. The canonical correlation analysis essentially jointly tests whether the correlations in the matrix below are simultaneously zero.BT, breath test kPCD; GE, proportion of gastric emptying at specified times. Open table in a new tab NOTE. All correlations are P < .001 except for correlation between GE 45 and BT 180, which was .004. Note that, in general, the largest correlation coefficients for GE scintigraphy and BT occur at the times when sampling occurred (eg, 45, 90, 120 minutes). However, later in the study (150 and 180 minutes), the correlations are not necessarily matched for time. The canonical correlation analysis essentially jointly tests whether the correlations in the matrix below are simultaneously zero. BT, breath test kPCD; GE, proportion of gastric emptying at specified times. The results of the multiple linear regression models (incorporating gender and BMI as covariates) provided predicted values for the GE proportions at 45, 90, 120, 150, 180, and 240 minutes. The predicted GE proportions are given by the equations in Supplemental Table 2.Supplementary Table 2Predicted GE Proportions at Different Time Points Based on Sex, BMI, and 13CO2 Excretion at Specified TimesGE45 = −0.00089 − 0.02941*Female + 0.00471*BMI + 0.01240*BT45 + 0.00158*BT90 − 0.00006*BT120 + 0.00023*BT150 − 0.00248*BT180GE90 = 0.01612 − 0.02309*Female + 0.00475*BMI + 0.00346*BT45 + 0.00650*BT90 + 0.00694*BT120 − 0.00033*BT150 − 0.00384*BT180GE120 = 0.03945 − 0.04090*Female + 0.00559*BMI + 0.00429*BT45 − 0.00167*BT90 + 0.01138*BT120 + 0.00319*BT150 − 0.00371*BT180GE150 = 0.07400 − 0.03407*Female + 0.00605*BMI + 0.00651*BT45 − 0.00449*BT90 + 0.00499*BT120 + 0.01062*BT150 − 0.00313*BT180GE180 = 0.12177 − 0.03421*Female + 0.00598*BMI + 0.00776*BT45 − 0.00564*BT90 + 0.00561*BT120 + 0.00232*BT150 + 0.00539*BT180GE240 = 0.37208 − 0.00923*Female + 0.00400*BMI + 0.00439*BT45 − 0.00433*BT90 + 0.00562*BT120 − 0.00392*BT150 + 0.00912*BT180 Open table in a new tab By using the predicted GE proportions from the BT measurements for each subject, a GE t1/2 value was computed for each subject (linear interpolation). The observed t1/2 values obtained in each individual by scintigraphy are plotted against the BT-estimated t1/2 values in Figure 3; the concordance was 0.86. The corresponding Bland-Altman plot is shown in Figure 4. Although the individual differences (scintigraphic t1/2 versus BT t1/2) varied from −53 to +75 minutes, the mean difference was less than 2 minutes. More generally, the mean values of the t1/2 values by BT and by scintigraphy over any random sample subset of 25 of the 129 patients wer
Referência(s)