Page 23 - Read Online
P. 23
Ladeira et al. Microbiome Res Rep 2023;2:9 https://dx.doi.org/10.20517/mrr.2023.01 Page 5 of 15
Table 1. Datasets used in the study
No antibiotic intake
All individuals No antibiotic intake
and adults only
Characteristic N= 9.515 N= 5.728 N= 4.921
Antibiotic intake 216 (3.6%)
Unknown 3.571
No 5.728 (100%) 4.921 (100%)
Health status
Control 7.016 (79%) 4.876 (85%) 4.348 (88%)
Adenoma 153 (1.7%) 39 (0.7%) 29 (0.6%)
Colorectal 439 (4.9%) 110 (1.9%) 74 (1.5%)
Metabolic 656 (7.4%) 495 (8.6%) 320 (6.5%)
Bowel 535 (6.0%) 99 (1.7%) 45 (0.9%)
Arthritis 94 (1.1%) 89 (1.6%) 85 (1.7%)
Behçet’s disease (BD) 20 (0.2%) 20 (0.3%) 20 (0.4%)
Unknown 602
Age category
Newborn 278 (2.9%) 137 (2.4%)
Child 322 (3.4%) 160 (2.8%)
School-age 135 (1.4%) 88 (1.5%)
Adult 7.745 (81%) 4.921 (86%) 4.921 (100%)
Senior 1,035 (11%) 422 (7.4%)
Westernized lifestyle 8.701 (91%) 5.577 (97%) 4.774 (97%)
n (%)
Given the multiple differences in analytical procedures between the studies included in the cMD, we first
investigated whether our analysis of the cMD database could reproduce published findings for
Bifidobacterium in humans, such as differences according to age, lifestyle, antibiotic use, and health status.
We found that subjects with a westernized lifestyle had higher relative abundances of Bifidobacterium
(Mann-Whitney, P < 0.001) [Supplementary Figure 2] and a higher prevalence of B. animalis and B. longum
[Supplementary Figure 3]. The gut microbiome of newborns was more enriched in Bifidobacterium (median
of 7.76 %, IQR 0.48%-41.3%) than that of the other age categories (median 2%, IQR 0.01%-9.8%) (Kruskal
Wallis, P < 0.001) [Supplementary Figure 2], with a higher prevalence of B. breve in children under the age
of three years (50%), decreasing to < 10% thereafter. B. adolescentis was more prevalent in adults (71%) than
in younger subjects (35%-50%) and seniors (60%). By contrast, B. longum was highly prevalent at all ages (>
80%) [Supplementary Figure 3]. The relative abundance of Bifidobacterium was significantly lower in most
of the disease groups, especially Bowel (IBD) (median 0.03%, IQR 0%-0.2%) than in healthy individuals
(median 3.4%, IQR 0.6%-10.7%) (Kruskal Wallis, P < 0.001) [Supplementary Figure 2]. In particular, B.
adolescentis was more prevalent in healthy individuals (70%) than in those with the disease, particularly for
IBD and metabolic diseases (40%-50%), whereas the opposite pattern was found for B. dentium
[Supplementary Figure 3]. Finally, the relative abundance of Bifidobacterium was lower in subjects with
recent antibiotic intake (Mann-Whitney, P < 0.001) [Supplementary Figure 2]. Despite the considerable
analytical differences between studies, this dataset reproduced the major associations previously reported
between Bifidobacterium and age , lifestyle , health status [1,29,30] , and antibiotic use [31-33] in individual
[28]
[27]
cohorts.

