Role of the gut microbiome in mediating lactose intolerance symptoms

example, in a highly cited large multicentre study (n=2600 patients), the CTC sensitivity for detection of ≥10 mm colonic polyp was found to be 90%. Another example is a pooled analysis (n=5328 patients) used recently by the US Preventive Service Task Force concluding that the CTC sensitivities to detect 6 mm and 10 mm colonic polyps were 86% and 89%, respectively, which are significantly higher than the results reported by Cash et al. Despite such disparate findings, there is no attempt to explain the differing results of this trial compared with prior larger validated trials. ► The manuscript states that ‘All CCE videos and CTC images were evaluated by one of two expert central readers who documented study quality as well as location and size of any colorectal polyps or masses’. It would be helpful to readers to understand the definition of ‘expert readers’ in terms of the number of CTC previously interpreted, degree of formal training and/or years of experience, especially in light of the surprisingly lower sensitivities of CTC described. ► Another missing component of this paper is the description of study limitations, even potential ones, which are an essential component of scientific discussion. ► The following are the striking differences between CCE and CTC: – Difference in examination duration: the median time needed for CCE was about 5 hours versus usually less than 1 hour for CTC. – Difference in intensity of bowel cleansing regimen: more intense bowel prep and cathartic boosters are needed for CCE. 6 – Difference in study completion rate: four patients assigned to CCE while one patient assigned to CTC had incomplete examination (the rate of incomplete examination is four times higher in the CCE arm compared with the CTC arm). Although these differences may not be directly related to diagnostic accuracy, a full discussion of both the advantages and the limitations of each screening modality would be helpful to readers when applying the results of any study to an individual patient’s circumstances. ► The wide exclusion criteria specific to the use of CCE (excluding patients with dysphagia, swallowing disorder, gastroparesis, bowel dysmotility, bowel stricture or fistula, patients with cardiac pacemaker or renal impairment) may limit the generalisability of the study findings. We thank you and the readers for the time and attention to our concerns. We believe that more studies are needed to reach a conclusive recommendation.

example, in a highly cited large multicentre study (n=2600 patients), the CTC sensitivity for detection of ≥10 mm colonic polyp was found to be 90%. 3 Another example is a pooled analysis (n=5328 patients) used recently by the US Preventive Service Task Force concluding that the CTC sensitivities to detect 6 mm and 10 mm colonic polyps were 86% and 89%, respectively, which are significantly higher than the results reported by Cash et al. 4 Despite such disparate findings, there is no attempt to explain the differing results of this trial compared with prior larger validated trials. ► The manuscript states that 'All CCE videos and CTC images were evaluated by one of two expert central readers who documented study quality as well as location and size of any colorectal polyps or masses'. It would be helpful to readers to understand the definition of 'expert readers' in terms of the number of CTC previously interpreted, degree of formal training and/or years of experience, especially in light of the surprisingly lower sensitivities of CTC described. ► Another missing component of this paper is the description of study limitations, even potential ones, which are an essential component of scientific discussion. 5 ► The following are the striking differences between CCE and CTC: -Difference in examination duration: the median time needed for CCE was about 5 hours versus usually less than 1 hour for CTC. 1 -Difference in intensity of bowel cleansing regimen: more intense bowel prep and cathartic boosters are needed for CCE. 1 6 -Difference in study completion rate: four patients assigned to CCE while one patient assigned to CTC had incomplete examination (the rate of incomplete examination is four times higher in the CCE arm compared with the CTC arm). 1 Although these differences may not be directly related to diagnostic accuracy, a full discussion of both the advantages and the limitations of each screening modality would be helpful to readers when applying the results of any study to an individual patient's circumstances. ► The wide exclusion criteria specific to the use of CCE (excluding patients with dysphagia, swallowing disorder, gastroparesis, bowel dysmotility, bowel stricture or fistula, patients with cardiac pacemaker or renal impairment) may limit the generalisability of the study findings.
We thank you and the readers for the time and attention to our concerns. We believe that more studies are needed to reach a conclusive recommendation.

Role of the gut microbiome in mediating lactose intolerance symptoms
Misselwitz et al recently reviewed the multifactorial aspects of lactose intolerance (LI). 1 Their work highlights the known effects of genetic makeup and dietary patterns in the occurrence of gastrointestinal symptoms in LI individuals. The authors define LI as the occurrence of gastrointestinal complaints on lactose exposure and discuss the wide variety of symptoms among LI patients. 1 Regarding lactose metabolism, Misselwitz et al mention the influence of the gut microbiome, in particular that Bifidobacterium or other lactose-fermenting bacteria are reported to affect the levels of lactose in the gut. However, the impact of the gut microbiome on the occurrence of gut-related LI symptoms remains unclear.
We previously observed that the Bifidobacterium abundance in the adult human gut depends on the interaction between LI genetic variants and dairy intake. 2 This observation complements other findings indicating a mutual relationship between the gut microbiome and host lactose metabolism. [3][4][5][6][7][8] However, these earlier analyses did not consider the occurrence of gastrointestinal symptoms. This inspired us to investigate whether the interplay of dairy consumption and Bifidobacterium affects the occurrence of gastrointestinal complaints in LI individuals.
We analysed data on gut complaints, genetics, gut microbiome and diet in 959 participants of the Lifelines-DEEP Dutch population cohort. We classified LI genetic status based on the presence of the functional variant rs4988235 associated with LI in Caucasian populations, as described earlier, 2 and we found the LI recessive genotype (G/G) in 81 individuals (8.4% of the cohort). We investigated microbiome composition using shotgun metagenomic sequencing of faecal samples 2 and selected the relative abundance of the Bifidobacterium genus for our analysis. Dietary intake, including milk and other dairy products, was assessed using a food frequency  questionnaire. Total dairy consumption, in grams per day, was retrieved using the Dutch Food Composition tables. 9 Notably, there were no reports of complete avoidance of dairy consumption among patients with LI. Lastly, gut complaints were assessed via a 7-day questionnaire where participants were asked to rank (from 1 to 5) their daily level of gastrointestinal discomfort in six categories: 'abdominal discomfort', 'bloating', 'burping', 'abdominal pain', 'flatulence' and 'nausea'. The sum of the mean daily scores from each category was used as the overall score of gastrointestinal complaints.
As reported previously, an increased Bifidobacterium abundance was observed in LI individuals compared with non-LI individuals (p Wilcox =4.56×10 -9 ) ( figure 1A) and was positively correlated with dairy intake in the LI group (R=0.22, p=0.05) but not in the non-LI group (R=0.02, p=0.48). No significant difference in dairy consumption was found between the LI and non-LI groups (p Wilcox =0.31). In LI individuals, we observed a positive correlation between Bifidobacterium abundance and total gut complaints score (R=0.33, p=0.003) (figure 1B). Of the six specific gut complaints, Bifidobacterium abundance was positively correlated with abdominal pain, discomfort and bloating (figure 1C-E) but not with any other complaints (all p>0.26). We then performed a mediation analysis to address how dairy intake and Bifidobacterium abundance would influence the occurrence of symptoms in LI individuals. Interestingly, we found that the association between dairy intake and gastrointestinal complaints was partially mediated by Bifidobacterium abundance (Prop mediated =43%, p=0.054) (figure 2).
Our results provide evidence that specific gut symptoms, experienced by LI patients, might be the result of Bifidobacterium abundance in the gut, rather than a direct effect of lactose intake.
This work supports initial reports where metabolic products of lactosefermenting bacteria may be related to LI symptom occurrence. 7