Brain and gut microbiota disorders in the psychopathology of anorexia nervosa

Causal effects of gut microbiota, metabolites, immune cells, liposomes, and inflammatory proteins on anorexia nervosa: A mediation joint multi-omics Mendelian randomization analysis

BACKGROUND

Anorexia nervosa (AN) is a significant psychological disorder influenced by environmental and genetic elements. Emerging research highlights the pivotal role of the gut microbiome in the development of diverse mental health conditions. This study aims to explore the causal effects and interactions of the gut microbiome, metabolites, immune cells, lipids, and inflammatory proteins on the risk of anorexia nervosa through mediation and multi-omics Mendelian Randomization (MR) analysis.

METHODS

This study used data from the FinnGen genome-wide association study (GWAS) of AN (N = 402,625), integrated with GWAS data on 473 of gut microbiota (N = 5959), 233 metabolites (N = 136,016), 731 immune cells (N = 3757), 179 lipids (N = 7174), and 91 inflammatory proteins (N = 14,824). This study used the univariate MR (UVMR), mediation MR analysis, and sensitivity analysis to assess the potential causal associations between these biomarkers and AN.

RESULTS

The inverse variance weighted (IVW) results suggest that 25 gut microbiota have causal effects on AN. Firmicutes E (OR: 0.294, 95 % CI: 0.107-0.806, P = 0.017), RUG147 (OR: 0.386, 95 % CI: 0.151-0.990, P = 0.048), CAG-977 (OR: 0.562, 95 % CI: 0.378-0.837, P = 0.005), Desulfobacterota A (OR: 0.651, 95 % CI: 0.466-0.909, P = 0.012), CAG-269 sp002372935 (OR: 0.673, 95 % CI: 0.483-0.937, P = 0.019), Klebsiella (OR: 0.684, 95 % CI: 0.566-0.827, P = 0.00009), Desulfovibrionia (OR: 0.706, 95 % CI: 0.538-0.926, P = 0.012), Klebsiella pneumoniae (OR: 0.737, 95 % CI: 0.600-0.906, P = 0.004), Desulfovibrionales (OR: 0.786, 95 % CI: 0.631-0.979, P = 0.031), CAG-776 (OR: 0.787, 95 % CI: 0.632-0.980, P = 0.032), Desulfovibrionaceae (OR: 0.788, 95 % CI: 0.635-0.978, P = 0.030). 13 gut microbiota were risk factors for AN, including Parachlamydiales (OR: 3.134，95%CI: 1.185-8.287, P = 0.021), Paenibacillus J (OR: 2.366，95%CI: 1.305-4.29, P = 0.005), Gillisia (OR: 1.947，95%CI: 1.135-3.339, P = 0.016), UBA1191 (OR: 1.856，95%CI: 1.221-2.822, P = 0.004), UBA7703 (OR: 1.843，95%CI: 1.032-3.289, P = 0.039), Faecalicatena sp002161355 (OR: 1.788，95%CI: 1.114-2.870, P = 0.016), Johnsonella ignava (OR: 1.742，95%CI: 1.031-2.944, P = 0.038), Staphylococcus aureus (OR: 1.614, 95%CI: 1.007-2.588, P = 0.047), Comamonas (OR: 1.522，95%CI: 1.004-2.307, P = 0.048), Ruminococcus D (OR: 1.24，95%CI: 1.050-1.464, P = 0.011), CAG-349 (OR: 1.198，95%CI: 1.048-1.370, P = 0.008), Ruminococcus D bicirculans (OR: 1.175，95%CI: 1.001-1.379, P = 0.048), CAG-177 (OR: 1.272，95%CI: 1.077-1.503, P = 0.005). Reverse MR analysis showed that causal effect of AN on 18 gut microbiota, but to a lesser extent. 12 metabolites have causal effects on AN. There are 7 protective factors, including glucose levels (OR: 0.700, 95%CI: 0.550-0.893, P = 0.004), isoleucine levels (OR: 0.769, 95%CI: 0.602-0.983, P = 0.036), phospholipids in large VLDL (OR: 0.856, 95%CI: 0.736-0.996, P = 0.044), total lipids in large VLDL (OR: 0.860, 95%CI: 0.740-0.999, P = 0.049), total lipids in small VLDL (OR: 0.863, 95%CI: 0.751-0.992, P = 0.038), free cholesterol in small VLDL (OR: 0.86, 95%CI: 0.752-0.996, P = 0.044), and free cholesterol in medium VLDL (OR: 0.866, 95%CI: 0.752-0.998, P = 0.047). There are 5 risk factors, including estimated degree of unsaturation (OR: 1.174, 95%CI: 1.009-1.367, P = 0.039), free cholesterol to total lipids ratio in small VLDL (OR: 1.199, 95%CI: 1.017-1.414, P = 0.031), phospholipids to total lipids ratio in small VLDL (OR: 1.216, 95%CI: 1.008-1.467, P = 0.041), total cholesterol levels in small HDL (OR: 1.241, 95%CI: 1.008-1.530, P = 0.042), and phospholipids to total lipids ratio in medium VLDL (OR: 1.280, 95%CI: 1.055-1.553, P = 0.012). Reverse MR analysis showed that AN had a causal effect on 15 metabolites. Mediation analysis reveals that the estimated degree of unsaturation mediates 0.69 % of the effect of Klebsiella pneumoniae on AN. Total lipids in small VLDL mediate 0.358 % of the effect of CAG-177 on AN, with a mediated proportion of 1.490 %. The mediation proportions for Estimated degree of unsaturation and Total lipids in small VLDL are relatively small. 36 immune cells have causal effects on AN. There are 7 protective factors, including Switched memory B cells %B cell (OR: 0.892，95%CI: 0.801-0.994, P = 0.038), CD127-CD8+ T cell absolute count (OR: 0.888，95%CI: 0.789-1.000, P = 0.049), IgD + CD24- B cell (OR: 0.917，95%CI: 0.862-0.975, P = 0.006), HVEM+ T cell (OR: 0.945，95%CI: 0.894-0.999, P = 0.045), CD40 + CD14 + CD16- monocyte (OR: 0.937，95%CI: 0.882-0.996, P = 0.038), CD64 + CD14 + CD16- monocyte (OR: 0.966，95%CI: 0.939-0.993, P = 0.016), CD8+ natural killer T cells (OR: 0.911，95%CI: 0.836-0.992, P = 0.032), HLA-DR+ T cells (OR: 0.921，95%CI: 0.866-0.980, P = 0.010), CD28-CD8+ T cells (OR: 0.886，95%CI: 0.792-0.991, P = 0.034). There are 26 risk factors. Reverse MR analysis showed that AN had a causal effect on 31 immune cells. AN increases the expression levels of five types of immune cells, including CD40 + CD14-CD16+ monocytes (OR: 1.087，95%CI: 1.004-1.177, P = 0.041), PDL-1+ CD14-CD16+ monocytes (OR: 1.082，95%CI: 1.002-1.168, P = 0.046), CD45+ CD33dim HLA-DR+ cells (OR: 1.145，95%CI: 1.019-1.287, P = 0.023), CD45+ basophils (OR: 1.164，95%CI: 1.036-1.307, P = 0.011), CD8+ natural killer T cells (OR: 1.102, 95%CI: 1.015-1.196, P = 0.020), and also decreases the expression levels of 26 immune cells. 6 liposomes showed exhibit protective effects against AN, including phosphatidylcholine (18:0_20:3) levels (OR: 0.852, 95%CI: 0.740-0.981, P = 0.026), phosphatidylcholine (O-18:2_18:1) levels (OR: 0.800, 95%CI: 0.672-0.952, P = 0.012), phosphatidylinositol (18:0_18:1) levels (OR: 0.873, 95%CI: 0.773-0.986, P = 0.029), phosphatidylinositol (18:1_18:2) levels (OR: 0.844, 95%CI: 0.734-0.971, P = 0.018), sphingomyelin (d38:1) levels (OR: 0.903，95%CI: 0.820-0.995, P = 0.039), and triacylglycerol (56:4) levels (OR: 0.786, 95%CI: 0.660-0.936, P = 0.007). There are 3 risk factors, including diacylglycerol (16:1_18:1) levels (OR: 1.208, 95%CI: 1.040-1.404, P = 0.014), phosphatidylcholine (18:1_18:3) levels (OR: 1.237, 95%CI: 1.003-1.526, P = 0.047), and phosphatidylinositol (16:0_20:4) levels (OR: 1.148, 95%CI: 1.003-1.314, P = 0.045). Reverse MR analysis showed that AN had a causal effect on 3 phosphatidylcholine (15:0_18:2) levels (OR: 1.075, 95%CI: 1.001-1.154, P = 0.048), phosphatidylcholine (O-16:2_18:0) levels (OR: 1.078, 95%CI: 1.002-1.159, P = 0.043), and triacylglycerol (51:1) levels (OR: 0.919, 95%CI: 0.850-0.994, P = 0.035). 6 inflammatory proteins have causal effects on AN, with protective factors including Glial cell line-derived neurotrophic factor levels (OR: 0.822，95%CI: 0.692-0.978, P = 0.027) and Interleukin-15 receptor subunit alpha levels (OR: 0.886, 95%CI: 0.789-0.995, P = 0.041) and risk factors including CC motif chemokine 4 levels (OR: 1.126, 95%CI: 1.011-1.254, P = 0.031), Interleukin-12 subunit beta levels (OR: 1.135, 95%CI: 1.033-1.248, P = 0.008), Monocyte chemoattractant protein-1 levels (OR: 1.152, 95%CI: 1.010-1.314, P = 0.035), and Sulfotransferase 1A1 levels (OR: 1.166, 95%CI: 1.006-1.351, P = 0.042). Reverse MR analysis showed that AN had a causal effect on Transforming growth factor-alpha (OR: 1.054，95%CI: 1.010-1.101, P = 0.016).

CONCLUSIONS

This study used large-scale and novel GWAS data, for the first time reveals through mediation analysis and multi-omics MR analysis the roles of gut microbiota, metabolites, immune cells, lipids, and inflammatory proteins in the pathogenesis of AN. These findings provide new biomarkers and targets for further prevention and treatment strategies.

[Consenso del grupo de trabajo de los trastornos de la conducta alimentaria de SENPE (GTTCA-SENPE). Evaluación y tratamiento médico-nutricional en la anorexia nerviosa. Actualización 2023]

Anorexia nervosa (AN) is a multifactorial disorder. A possible role of the social network and the gut microbiota in pathogenesis has been added. Exogenous shocks such as the COVID19 pandemic have had a negative impact on patients with AN. The potential medical and nutritional impact of malnutrition and/or compensatory behaviors gives rise to a complex disease with a wide range of severity, the management of which requires a multidisciplinary team with a high level of subject matter expertise. Coordination between levels of care is necessary as well as understanding how to transition the patient from pediatric to adult care is essential. A proper clinical evaluation can detect possible complications, as well as establish the organic risk of the patient. This allows caregivers to tailor the medical-nutritional treatment for each patient. Reestablishing adequate nutritional behaviors is a fundamental pillar of treatment in AN. The design of a personalized nutritional treatment and education program is necessary for this purpose. Depending on the clinical severity, artificial nutrition may be necessary. Although the decision regarding the level of care necessary at diagnosis or during follow-up depends on a number of factors (awareness of the disease, medical stability, complications, suicidal risk, outpatient treatment failure, psychosocial context, etc.), outpatient treatment is the most frequent and most preferred choice. However, more intensive care (total or partial hospitalization) may be necessary in certain cases. In severely malnourished patients, the appearance of refeeding syndrome should be prevented during renourishment. The presence of AN in certain situations (pregnancy, vegetarianism, type 1 diabetes mellitus) requires specific care. Physical activity in these patients must also be addressed correctly.

Longitudinal analysis of the gut microbiome in adolescent patients with anorexia nervosa: microbiome-related factors associated with clinical outcome

There is mounting evidence regarding the role of gut microbiota in anorexia nervosa (AN). Previous studies have reported that patients with AN show dysbiosis compared to healthy controls (HCs); however, the underlying mechanisms are unclear, and data on influencing factors and longitudinal course of microbiome changes are scarce. Here, we present longitudinal data of 57 adolescent inpatients diagnosed with AN at up to nine time points (including a 1-year follow-up examination) and compare these to up to six time points in 34 HCs. 16S rRNA gene sequencing was used to investigate the microbiome composition of fecal samples, and data on food intake, weight change, hormonal recovery (leptin levels), and clinical outcomes were recorded. Differences in microbiome composition compared to HCs were greatest during acute starvation and in the low-weight group, while diminishing with weight gain and especially weight recovery at the 1-year follow-up. Illness duration and prior weight loss were strongly associated with microbiome composition at hospital admission, whereas microbial changes during treatment were associated with kilocalories consumed, weight gain, and hormonal recovery. The microbiome at admission was prognostic for hospital readmission, and a higher abundance of Sutterella was associated with a higher body weight at the 1-year follow-up. Identifying these clinically important factors further underlines the potential relevance of gut microbial changes and may help elucidate the underlying pathophysiology of gut-brain interactions in AN. The characterization of prognostically relevant taxa could be useful to stratify patients at admission and to potentially identify candidate taxa for future supplementation studies aimed at improving AN treatment.

The Role of the Gut Microbiota in Anorexia Nervosa in Children and Adults-Systematic Review

Among the factors incriminated in the appearance of eating disorders, intestinal microbiota has recently been implicated. Now there is evidence that the composition of gut microbiota is different in anorexia nervosa. We gathered many surveys on the changes in the profile of gut microbiota in patients with anorexia nervosa. This review comprehensively examines the contemporary experimental evidence concerning the bidirectional communication between gut microbiota and the brain. Drawing from recent breakthroughs in this area of research, we propose that the gut microbiota significantly contributes to the intricate interplay between the body and the brain, thereby contributing to overall healthy homeostasis while concurrently impacting disease risk, including anxiety and mood disorders. Particular attention is devoted to elucidating the structure and functional relevance of the gut microbiota in the context of Anorexia Nervosa.