Detection capacity of small intestine bacterial or methanogen overgrowth by lactose and fructose breath testing in the adult population

Introduction

Small intestine bacterial overgrowth (SIBO) is a functional bowel disorder characterized by excessive proliferation of bacterial microbiome in the small intestine. Manifestations of this entity include bloating, flatulence, abdominal pain and diarrhea or constipation [1, 2]. These symptoms can be due to abnormal intestinal permeability and motility, a higher level of inflammation or immune activation, malabsorption of nutrients or fermentation of substrates in the small bowel [3].

In some cases, this overgrowth may be caused by intestinal methanogen overgrowth (IMO) [4]. These methanogenic organisms, primarily Methanobrevibacter smithii [5] are not bacteria but archaea. Hence a different term has been suggested to refer to their overgrowth [1]. IMO has been associated with constipation, due to the inhibitory effect of methane (CH₄) on intestinal peristalsis [6]. A large retrospective study demonstrated that IMO is a different entity from SIBO and that excessive CH₄ production is more frequent in elderly patients. The study also revealed that malabsorption of nutrients such as vitamin B12 is less frequent in IMO [7]. Additionally, a different antibiotic treatment is indicated for IMO. The combination of neomycin and rifaximin has been proven to be more effective in IMO patients than rifaximin alone, the treatment indicated for SIBO [8, 9].

The gold standard diagnostic study for SIBO is jejunal aspirate culture. However, its use is challenging, as it is an expensive invasive test and a standard cut-off of colony-forming units per ml has not yet been established [3]. Hence, the most widely used study for testing suspected SIBO or IMO is exhaled breath testing (BT). BT is a low priced, non-invasive, easy-to-use test that is widely available in a large number of laboratories. This test measures hydrogen (H₂), methane (CH₄), and carbon dioxide (CO₂), with the latter being used as a correction factor and to verify correct sample collection.

A variety of guidelines and consensus documents are available for the correct performance of BTs, including preparation instructions for patients undergoing a BT. The most relevant guidelines are the American [10] and the European [11]. However, these guidelines use different cut off points for the interpretation of SIBO-BT. Whereas the American guideline establish a rise in H2 of ≥20 parts per million (ppm) from baseline at 90 min, the European guideline only establish early elevation of H₂≥10–12 ppm from baseline, without providing any time interval. The American guideline further suggest that CH₄ elevation (≥10 ppm) indicates the presence of IMO, whereas this aspect is not mentioned in the European guideline. The most widely recommended substrate is 50–75 g of glucose or 10–20 g of lactulose. Of note, another substrate similar to lactulose is used in Spain, lactitol. These substrates are widely used for diagnosis of SIBO, as they both are nonabsorbable sugars. A recent meta-analysis revealed a slightly higher diagnostic performance for glucose, as compared to lactulose [12].

As any other test, BTs also have some limitations such as it being an indirect measurement method. In addition, accelerated/reduced orocecal transit (due to physiological factors, previous surgery, use of drugs, and/or diseases) can interfere with results. Other limitations include its lack of correlation with the results of small bowel aspirate culture [13]. In addition, since methanogenic archaea use H₂ as a substrate for CH₄ production, the levels of H₂ are usually low in the presence of IMO; therefore, it is critical that the instruments used to measure exhaled breath gases in BTs also measure CH₄ concentration [14].

On another note, malabsorption/intolerance to carbohydrates such as fructose (FI) or lactose (LI) can be related to the presence of SIBO and mimic its symptoms, as SIBO can reduce their absorption due to abnormal villi function and/or destruction of intestinal mucosal enzymes [15, 16]. Thus, FI/LI-BTs are frequently requested for patients with suspected SIBO/IMO. FI/LI-BTs may provide evidence suggestive of SIBO or IMO in the form of early H₂ elevation or high CH₄ concentrations [17].

Although the American consensus guideline recommend SIBO-BTs to be performed prior to carbohydrate malabsorption/intolerance BTs [10], for practical reasons, and due to the non specificity of symptoms, several BTs are often requested in the same medical consultation. Understanding the concordance between FI/LI-BTs and SIBO-BTs would be useful to avoid repeated BTs, recurrent consultations and delayed diagnosis, thereby reducing costs. However, little attention has been paid to this issue, with most studies having focused on positive cases for FI/LI that could be secondary to SIBO [18, 19].

The objective of this study was to examine whether findings suggestive of SIBO and IMO on FI/LI-BTs in the adult population were consistent with the results of SIBO-BTs. This study will extend knowledge on the sensitivity and specificity of FI/LI-BTs in detecting SIBO and their concordance in identifying IMO.

Materials and methods

Results

Demographic data and characteristics of BTs

Of the 652 BTs performed at our hospital during the study period, 146 were eligible and corresponded to 67 SIBO-BTs, 47 FI-BTs, and 32 LI-BTs. These BTs were performed in 67 patients, of whom 82 % were female, with a mean age of 47.8 ± 16.9 years (range 17.7–85.6).

Figure 1 details the types of BTs performed and the time intervals between tests. The majority of patients (52 %) underwent a FI-BT and a SIBO-BT, whereas only 18 % underwent the three types of BTs. The most frequent time interval between BTs was 1 week (58 %), followed by 2 weeks (24 %) and 3–6 weeks (18 %).

Figure 1:

Characteristics of exhaled breath tests (BTs) included in the final analysis according to the type of BT: fructose intolerance (FI); lactose intolerance (LI); or small intestine bacterial overgrowth (SIBO), and time period between the first and the last BT performed in the same patient.

In total, 52.9 % of patients with a positive FI-BT also showed intolerance, with this percentage rising to 77.8 % in patients with a positive LI-BT.

Discussion

BTs are a useful tool to evaluate functional disorders in patients with non-specific gastrointestinal symptoms such as bloating, abdominal pain or flatulence, due to their low cost and simplicity [11]. A study in 1,230 patients in whom underlying gastrointestinal disease had been excluded endoscopically and radiologically, revealed that 45 % had SIBO, FI and/or LI, which were detected by BTs [22].

This study contributes relevant evidence to the limited literature available on findings suggestive of SIBO on FI/LI-BTs. Additionally, this is the first study to provide data on the concordance of these tests in detecting the presence of IMO. The rate of false positives for SIBO on FI-BTs was 27 %. False positives were defined as early H₂ elevations suggestive of SIBO, which were not confirmed by SIBO-BT with lactitol. In contrast, the rate of false positives on LI-BTs was as low as 6 %, which indicated a high specificity. LI-BT also provided optimal results for the detection of IMO, with a level of concordance of 100 %. In contrast, the rate of false negatives for FI-BTs was 27 %.

IMO can be currently identified with gas chromatographs with capacity to measure both CH₄ and H₂. The prevalence of IMO could be 30–60 % of all overgrowths when tested by BT [23]. In our center, IMO accounted for 57 % of all overgrowths detected using this method [24]. According to our experience, when lactitol is used as a substrate, IMO manifests as a plateau where CH₄ is elevated at baseline, with values ≥10 ppm at all points of the BT curve. In contrast, the American guideline, which are based on lactulose or glucose, defined a positive test for IMO as CH₄ ≥10 ppm at any point of the curve [10]. On the other hand, the European guideline did not include any recommendation for the detection of IMO [11]. With lactitol, CH₄ rises in parallel to H₂. As a result, CH₄ is frequently ≥10 ppm at 100 min or later. Our interpretation is consistent with that of a recent study that revealed a high sensitivity (86 %) and specificity (100 %) in detecting IMO in patients with baseline samples containing CH₄≥10 ppm, as measured by glucose and lactulose SIBO-BT [25].

This is not only the first study to evaluate the concordance between FI-/LI-BTs and SIBO-BTs in detecting IMO. In addition, to the best of our knowledge, this is the first study to evaluate in an accurate and precise way the performance of FI-/LI-BTs in detecting SIBO. A previous study already identified SIBO as a factor that increased the probability of LI in patients with chronic diarrhea, although a thorough comparative study was not performed [26]. Previous studies only revealed a higher percentage of positive results for LI or sorbitol intolerance – but not for FI – when patients were found to be positive on SIBO-BT [18]. Evidence has also been provided that SIBO was a frequent condition (75 %) causing LI solely detected by BTs but not by oral lactose tolerance test [19]. According to an Asian study, the prevalence of LI secondary to SIBO was 15 % [27], a slightly lower percentage to that observed in our study (29 %). Finally, a recent study reported a higher percentage of cases of SIBO (28 % vs. 7 % in controls) in patients with lactase deficiency confirmed by biopsy [28]. Our study provides novel data on FI suspected to be secondary to SIBO, with a prevalence of 70 %, or the lower rate of false positives (6 %) between LI-BTs and SIBO-BTs in relation to SIBO, as compared to FI-BTs (27 %).

The main strengths of this study are that it is based on BTs performed over a 3-year period using the same gas chromatograph and in patients in whom different BTs were carried out in a short period of time. A limitation of this study is that H₂ elevation may be secondary to an accelerated intestinal transit. However, this disorder would affect SIBO-BT and FI/LI-BTs results equally. Additionally, the use of lactitol and exhaled breath sampling every 25 min for the SIBO-BT (which hinders measuring gases at 90 min) do not fully align with the recommendations of the American or European guidelines [10, 11]. As it occurs in most centers, duodenal aspirate cultures were not performed in routine practice; therefore, SIBO-BT results could not be compared against those of this gold standard test. The results reported are applicable to comparative studies of FI/LI-BTs and SIBO-BTs when lactitol is used as substrate. Studies with a larger number of patients and using other substrates (glucose or lactulose) are needed to confirm our findings. The production of hydrogen sulfide was not measured either, and it could be the cause of the presence of flat curves without H₂ or CH₄ elevation in five patients, which were interpreted as negative for SIBO and IMO.

In conclusion, this study demonstrates that detection of SIBO or IMO in LI-BT has a high level of concordance with SIBO-BT results. This means that SIBO-BT would not be necessary, which would ultimately prevent delayed diagnosis and treatment. In contrast, FI-BT showed a lower level of concordance with SIBO-BT, as 27 % of cases of detected SIBO will be false positives, and IMO would not be identified in another 27 %. These results highlight the need for highly-trained professionals in the optimal interpretation of BTs. This added to an adequate indication would ensure correct diagnosis and effective treatment, jointly with improved cost-effectiveness of these tests.