Fig. 5.

Gut microbiome and craving. Regression models (lm() function) of CLR-transformed species counts were related to craving variables, adjusting for BMI, dietary intake and the number of days since the last BD episode. False Discovery Rate (FDR) using sequential modified Bonferroni correction for multiple testing was applied. The effect (β) is represented in red (increased) or blue (reduced) with higher colour intensity representing a bigger effect. Opaque points represent effects that pass FDR. Craving was associated with reductions in the Ruthenibacterium lactiformas species, both at baseline and at follow-up (3 months later). Gut-brain modules (the potential of the microbiome to produce chemicals with neuroactive potential) was calculated with R Gomixer tool. A number of gut-brain modules showed associations with craving, such as reduced butyrate and inositol synthesis and increased acetate, glutamate and tryptophan synthesis. Associations were found for craving at baseline and at follow-up (3 months later). Baseline craving dimensions showing significant associations were expectancy and purposefulness (first and second column respectively) and at follow-up obsessed and regulated dimensions (third and fourth column).