European Journal of Clinical Investigation (1 987) 17, 189-1 93
Seeking clues for a positive diagnosis of the irritable bowel syndrome
P. VERNIA, G. LATELLA, F. M. MAGLIOCCA, G. MANCUSO* & R. CAPRILLIT, Cattedre di Gastroenterologia, e tAquila and *Istituto di Pediatria, Unversita di Roma, Italy
Received 3 April 1986 and in revised form 27 November 1986
Abstract. Despite its high prevalence the irritable bowel syndrome (IBS) lacks acceptable pathophysio- logical markers and its diagnosis largely depends on the exclusion of underlying organic disease. Systemic acid-base balance, serum electrolytes and the compo- sition of faecal water (electrolytes and organic anions), were studied in thirty-eight diarrhoea1 patients out of a series of ninety-three consecutive IBS patients. Only patients with diarrhoea as the predominant symptom were included in the study to evaluate whether this subgroup could provide the clue for a positive diagno- sis of the syndrome. Serum electrolytes and systemic acid-base balance were within the normal range. Faecal electrolytes were also normal (Na 26.6 & 19.3 SD; K 66.8f28.3; C1 19.1 15.2 mEq 1-I), despite the finding of a moderately increased 24-h faecal output. The K:Na ratio was also within the normal range. These data are in agreement with the lack of systemic changes observed in IBS patients even with profuse or longstanding diarrhoea. Both faecal short chain fatty acids and lactic acid were increased in patients vs. controls, but a considerable overlap with normal values was observed (1 3 1.4 f 62.6 SD vs. 108.5 f 58.3 mEq 1 - I ) . Only lactic acid concentration was signifi- cantly higher than in controls (1.3 f 1.2 vs. 0.5 f 0.2). Despite these findings i t is concluded that the subgroup of IBS patients with diarrhoea also appears to lack a pathophysiological marker and does not provide clues for a positive diagnosis of this syndrome.
Keywords. Irritable bowel syndrome, faeces, fatty acids, volatile.
Introduction The irritable bowel syndrome (IBS) affects 15-30% of the otherwise normal population in western countries [ 1-31 and is the most common disease category seen by gastroenterologists in out-patients [4,5]. Despite the
Correspondence: P. Vernia. MD, Cattedrd di Gastroenterologia ( I ) , 2a Clinica Medica, Universitd di Roma. Policlinico Umberto I, 00100 Roma, Italy.
high prevalence, attempts to identify positive diagnos- tic features [6,7] have so far been unrewarding.
Some symptoms such as abdominal pain preceeding evacuation, abdominal distension, irregularity of bowel pattern, sense of incomplete evacuation and mucus in the stool have consistently been shown to be associated with IBS. However, when these symptoms and simple laboratory findings were evaluated to estimate their relative diagnostic value, the most heavily weighed items were found to be those pathog- nomonic for non-functional diseases . Thus, the diagnosis of IBS still largely depends on the exclusion of underlying organic disease .
The lack of acceptable pathophysiological markers and of a precise definition of the syndrome leads to the inclusion in IBS of patients with varying clinical presentation ranging from spastic colon, scybalous constipation and abdominal pain to atonic colon and painless diarrhoea. A more precise classification of the clinical subgroups within IBS may thus prove useful for understanding this syndrome better.
The present study was focused only on the subgroup of IBS patients in whom the major complaint was diarrhoea, as this is one of the most common symp- toms in IBS, affecting 20-35% of patients [7,10]. Furthermore, since diarrhoea might be expected to induce deviations from the norm in terms of systemic electrolyte and acid-base balance it might provide the clue for a positive diagnosis of IBS.
Patients and methods
Patients The investigation was carried out on thirty-eight
patients in whom the major complaint was diarrhoea, selected from a series of ninety-three IBS attending the out-patient department of the GI Unit of the Univer- sity of Rome, during the period 1982-1984. Diagnosis of IBS was based on a typical longstanding history of abdominal symptoms (diarrhoea, constipation or alternation of the two, hypersecretion of colonic mucus, abdominal pain, flatulence and general lack of subjective well being), normal finding of barium enema
190 P. VERNIA et al.
and upper GI series, pancreas and biliary tract echo- graphy, and fibersigmoidoscopy with biopsies. Nor- mal values for ESR, haemoglobin, WBC, ALT, AST, yGT, alkaline phosphatase, amylase, thyroid hor- mones, serum proteins, urine sediment and stool examination for pathogenic bacteria, parasites and occult blood were also included in the diagnostic criteria.
In this study, diarrhoea was defined as the presence of three or more, liquid or semi-liquid, bowel move- ments per day. Only those IBS patients presenting diarrhoea more than 50% of the time during the month prior to the study were included in the investigation. All patients were in a diarrhoea1 state at the time of the study. They were all on a free diet, which did not include fibre supplements. Patients with known lactose intolerance or who showed a clear benefit from a non- milk diet were not included in the present series. No patient in this series had severe postprandial watery diarrhoea, nor showed high faecal pH associated with a reversed faecal K:Na ratio [ I I], thus ruling out the presence of patients with idiopathic bile salt catharsis
Each patient was interviewed concerning the pres- ence or absence of abdominal pain, straining at stools, mucus in the stool, abdominal distension and flatu- lence, looser stools with pain onset, irregular bowel pattern or persistent diarrhoea.
Patients were instructed to collect faeces over a 24-h period in preweighed plastic containers. Thiomersal (15 ml, 1 : 10 000 solution) was added immediately after the first evacuation to minimize bacterial over- growth. Faeces were brought to the laboratory at the end of the 24-h period and processed immediately or frozen at - 12C for 1-5 days. Upon completion of the 24-h collection, arterial and venous blood samples were drawn in all patients in order to measure systemic acid-base balance and the serum electrolyte concen- trations. Informed consent was obtained in all cases.
Analytical methods Arterial blood pH and pCO2 were measured by
means of a blood gas analyser (Radiometer Mod. 25, Copenhagen, Denmark). Bicarbonate concentration was calculated using the Henderson-Hasselbach equa- tion.
Faecal samples were weighed, and faecal water extracted using an in-riitro dialysis method, as de- scribed elsewhere [ 131.
Sodium and potassium were analysed in blood and faecal water by flame photometry; chloride by poten- tiometric titration. Short chain fatty acids (SCFA) were extracted in ether from the faecal water of eighteen patients and twenty-nine normal controls using the procedure described by Holdeman et al. [I41 and analysed by means of a Varian Model 3700 thermal conductivity gas-liquid chromatograph employing Supelco (SP 1000, 1 % H3P04, 100-120 mesh packed) columns. Solution of volatile SCFA at
10 mEq 1 - ' were made from 90-98% pure commercial preparations (Gold Label, Aldrich Chemical Co., Milwaukee, WS) of the following acids: acetic, propio- nic, isobutyric, isovaleric, valeric and caproic. The extracts of the 10 mM solution ofeach SCFA submitted to the same procedure as faecal water samples were used as standards. D- and L-lactic acid were measured with an enzymatic method using specific L- and D- lactate dehydrogenase [ 151. The concentration of fae- cal water electrolytes in IBS patients was compared to that in fifty normal subjects. Control subjects were selected from healthy volunteers with normal bowel habits, who did not experience adverse abdominal symptoms. They were all on a free diet, which did not include fibre supplements. Sex incidence and age were comparable in the control and patient group (38.2 f 12 SD vs. 2 & 14 years; female : male ratio 1.8 and 2.1 in the control and in the patient group, respectively).
In order to exclude the possibility that some clues might become apparent only in some subgroups of patients, data from IBS patients were analysed in relation to the presence or absence of the individual symptoms sought during the interview, and of the 24-h faecal weight. Faeces of 225 g per day were considered as the upper normal limit being in excess of the mean value plus one standard deviation of the fifty control subjects.
Stat istical unalysis Data were analysed using the Mann-Whitney test.
The linear regression test was used for correlation analysis.
Systemic. acid-base balance and serum electrolytes The mean values for systemic acid-base balance
were within the normal range in the present series of patients (pH 7.41 f0.03 SD; pC0236.1 k 8.2 SD mmHg; HC03- 23.3 f 2.4 SD mEq 1 - I ) . Four patients showed slight respiratory alkalosis, which may be attributed to hyperventilation due to the arterial blood sampling. One patient showed mild metabolic acidosis
Table 1. Faecal water electrolytes in irritable bowel syn- drome and normal controls
24-h weight* Nd K CI K Nd
IBS 224 266 668 19 I 4 8 (11=38) +I31 0 t 5 1 9 3 i 2 8 3 + I 5 2 k5X P 0005 N S NS NS NS Controls I47 219 611 1 4 0 5 2 01-50) i 8 0 6 1 1 7 7 + 2 6 6 * I 0 4 i 4 8
~~ - ~- .. ~ -~
* Twenty-four hour faecal weight expressed in g 24 h - I
t Mean values+SD. electrolyte concentrations expressed in mEq I I.
DIAGNOSIS OF IRRITABLE BOWEL SYNDROME 191
Table 2. Faecal water short chain fatty acids in irritable bowel syndrome and normal controls
Total Acetic Propionic Isobutyric Butyric Isovaleric Valeric Total SCFA lactic acid
IBS 83.2' 23.7 2.3 18.6 1.7 2.0 131.4 1.3 (n=18) k43.71 i 1 3 . I 1.9 15.1 k0 .9 i 1 . 3 i62 .6 i 1 . 2 P NS NS NS NS NS NS NS 0.001 Controls 70 17.1 2.0 13.0 2.6 2.3 108.5 0.5 (n=29) k49.8 k8.0 +1.0 8 8 . 5 k1.8 k1 .2 k58.3 k0.2
* Concentrations expressed in mEq 1 - ' t Mean values + SD.
Table 3. Per cent incidence of individual short chain fatty acids
Caproic & Acetic Propionic Isobutyric Butyric Isovaleric Valeric isocaproic
I BS 64,6* 18.5 I .9 12.2 1.2 I .2 0.4 +1.7 i 7 . 2 k0 .6 k0 .6 k0 .6 - +4.7t k5.1 -
P NS NS NS NS NS NS NS Controls 64.5 16.0 1.9 12.2 2.4 2. I 0.3
54 .1 k 4 - 2 k1 .6 i 5 . 3 kO.5 +0.6 k0 .4
* Data expressed as per cent of total SCFA. t Mean values* SD
(pH 7.36, pCOz 39 mmHg; HC03- 21.7 mEq I - ' ) . All individual patients had normal serum electrolytes (mean: Na 140.5k2.9 SD; K 4.1 f0.4; CI 103.2f3.5 mEq I - ' ) .
Faecal analysis The 24-h faecal weight was moderately but signif-
cantly increased in IBS patients compared to that of normal subjects. The electrolyte component of faecal water and the stool K : Na ratio were within the normal range (Table 1).
The organic component of the faecal water of IBS patients was significantly different from that of con- trols (Table 2). Irritable bowel syndrome patients presented a slightly increased concentration of total
SCFA resulting from higher concentrations of the major (C2-G) SCFA. Caproic acid was present in low concentrations, ranging from trace amounts to 1 . I mEq 1 - I , in some 50% of the IBS patients and normal subjects. Traces of isocaproic acid were also present in about one-third of all faecal samples. The concentra- tions of the latter SCFA were similar in IBS and normal subjects.
Although slightly increased concentrations of SCFA were observed in IBS, the per cent incidence of any single acid was almost identical to that in controls
Table 5. Faecal water composition in irritable bowel syndrome patients with normal and increased 24-h
Table 4. Incidence of individual symp- toms in thirty-eight irritable bowel syn-
Symptom % present -.
Abdominal pain 76.3
Mucus in the stools 63. I Irregular bowel pattern* 52,6 Persistent diarrhoea 47.3
Straining at stools 15.8
Looser stool with pain onset 7 1.1
Abdominal gas distension 44.7
* With predominant diarrhoea at the time of the study.
Total Nd K CI SCFA
Normal weight? 19 5: 667 175 156 I (n = 22) * I 1 4 i 2 9 4 k 1 2 2 +484 P NS NS NS NS Increased weight 28 6 649 23 4 1070 ( n = 16) + I 8 9 + 2 0 8 rfr207 +722
* Total SCFA were measured in eleven patients with normal. and seven patients with increased, 24-h faecal weight.
? T h e upper normal limit was chosen as 227 g 24 h -'. being in excess of the mean value plus 1 SD of fifty normal controls.
1 Data expressed as mEq I & SD.
192 P. VERNIA et al.
(Table 3). Lactic acid concentrations, on the other hand, were significantly higher in IBS patients vs. controls (Table 2). The discrepancy was entirely due to an increase in D-lactic acid (1.0 0.9 vs. 0.3 k 0.2 mEq 1 - I ) whereas L-lactate concentration was almost iden- tical in the two groups (0.27 k0.26 vs. 0.21 k0.11 mEq I - ) .
Indiridual symptoms and laboratory data
Data concerning the incidence of the most common IBS symptoms are shown in Table 4. The presence or absence of any single symptom was not related to specific patterns of faecal data. Also classifying the patient population according to a normal ( 227 g day-) faecal weight did not help to determine any difference between sub- groups of IBS patients (Table 5).
Correlation analysis A significant inverse correlation was found between
the 24-h faecal weight and acetic acid concentration ( r = - 0.598; P < 0.05). An acceptable correlation coef- ficient was present ( r =0.491) between faecal weight and total SCFA.
Discussion The results of this study show that diarrhoea in IBS is consistently associated with normal concentration of serum electrolytes and normal acid-base balance. These data are in keeping with the observations that diarrhoea in IBS does not significantly affect systemic acid-base homeostasis [ 161. Although all patients complained of diarrhoea, and had three or more bowel movements per day, their faecal output was only moderately increased compared to that of normal controls (235 vs. 147 g day-). The concentrations of sodium, potassium and chloride in faecal water in IBS patients were almost identical to those in normal subjects. With only two exceptions, all IBS patients had a normal faecal K : Na ratio, indicating that the colon had maintained normal electrolyte exchange mechanisms despite the excretion of moderately increased amounts of faeces. These findings explain why prolonged diarrhoea does not modify the concen- trations of serum electrolytes and systemic acid-base balance.
The faecal concentration of all major SCFA was slightly increased in IBS but, on account of the wide scattering of data, this observation can not be usefully employed for clinical purposes. IBS patients and control subjects did not supplement their diet with non-absorbable carbohydrates. Therefore, th...