Menopause Factors and Alterations in Gut Microbiota and Insulin Homeostasis: A Cross-Sectional Analysis of the Microbiome and Insulin Longitudinal Evaluation Study (MILES)

Vincent J Maffei; Alain G Bertoni; Alexis C Wood; Jerome I Rotter; Osa Crago; Yii-Der I Chen; Joseph F Petrosino; Kristi L Hoffman; Mark O Goodarzi; Elizabeth T Jensen

doi:10.1111/dom.16578

. Author manuscript; available in PMC: 2026 Jul 30.

Published in final edited form as: Diabetes Obes Metab. 2025 Jul 30;27(10):5444–5454. doi: 10.1111/dom.16578

Menopause Factors and Alterations in Gut Microbiota and Insulin Homeostasis: A Cross-Sectional Analysis of the Microbiome and Insulin Longitudinal Evaluation Study (MILES)

Vincent J Maffei ¹, Alain G Bertoni ¹, Alexis C Wood ², Jerome I Rotter ³, Osa Crago ¹, Yii-Der I Chen ³, Joseph F Petrosino ⁴, Kristi L Hoffman ⁴, Mark O Goodarzi ^5,^*, Elizabeth T Jensen ^1,^*

PMCID: PMC12715721 NIHMSID: NIHMS2093530 PMID: 40735808

Abstract

Aim

To assess the risk for impaired insulin homeostasis as a function of menopause-related factors and gut microbiota dysbiosis in non-diabetic, post-menopausal women.

Materials and Methods

Baseline data (n=175 women) from the Microbiome and Insulin Longitudinal Evaluation Study (MILES) were used, including insulin and dysglycemia indices calculated from a 2-hour oral glucose tolerance test, untargeted peripheral metabolomics, targeted peripheral short chain fatty-acid levels, and fecal bacterial microbiota surveyed by whole-metagenomic sequencing.

Results

After adjustment for covariates, menopause age <50 years and use of hormone replacement therapy (HRT) were associated with lower Matsuda et. al. insulin sensitivity index levels (β=−0.232, CI=[−0.450,−0.014] and β=−0.275, CI=[−0.444,−0.107], respectively) but not pre-menopausal gynecologic surgery. Pre-menopausal gynecologic surgery was significantly associated with fecal microbiota beta-diversity driven by a relative increase in diabetogenic Ruminococcus gnavus and Clostridium species and a decrease in protective Alistipes species and Akkermansia muciniphila relative abundances. A reduction in two glycerophospholipids in the plasmalogen class significantly statistically mediated an inverse association between gynecologic surgery before menopause and insulin sensitivity.

Conclusions

Menopause age and history of HRT are more strongly associated with insulin resistance than gynecologic surgery before menopause. However, gynecologic surgery is associated with shifts in gut microbial composition and plasma metabolite levels with a potential to contribute to future diabetes risk.

Introduction

Hysterectomy without concomitant bilateral oophorectomy may contribute to ovarian function alterations that increase risk of earlier menopause^1–3 and development of chronic comorbidities, including type two diabetes mellitus (T2DM)^4–6. After bilateral oophorectomy, menopause occurs immediately, which may also confer risks for similar incident chronic disease, particularly for oophorectomy at <50 years of age⁷. Earlier, non-surgical menopause age has also been associated with T2DM⁸. Novel methodologies are needed to identify women at risk for incident T2DM and to prevent its acquisition after menopause.

Signatures of impaired insulin homeostasis, such as insulin insensitivity and oral glucose intolerance, precede development of T2DM⁹. The mechanisms underlying these changes after menopause remain incompletely explored. Hypoestrogenism may augment insulin resistance¹⁰, glucose intolerance¹¹, and systemic inflammation^12,13, which further compromises insulin sensitivity¹⁴. Hormone replacement therapy (HRT) for peri-menopause or surgical menopause hypoestrogen management appears to improve systemic inflammation¹⁵ and decrease T2DM risk¹⁶. However, this benefit is lost after HRT discontinuation¹⁷ as potential side-effects have limited long-term use¹⁸.

In pre-clinical models, hormone receptor signaling supports the normal functioning of the gastrointestinal mucosa, which may also be undermined in menopause¹³. A compromised gut barrier promotes host inflammation by facilitating exposure to enteric microbial antigens¹⁹. Gut dysbiosis has been described in the menopause transition²⁰ and T2DM²¹. In pre-clinical models of surgical menopause, ovariectomy induces acquisition of metabolic syndrome traits and loss of microbiota-derived byproducts, such as short chain fatty-acids (SCFA)²², that support metabolic homeostasis, suggesting a contributory role for dysbiosis in the broader etiology of acquired metabolic disorders. The role of gut dysbiosis on T2DM risk after menopause remains an underdeveloped area of research.

The objective of this study, derived from the Microbiome and Insulin Longitudinal Evaluation Study (MILES)²¹, was to investigate whether menopause-related factors, including hysterectomy and/or bilateral oophorectomy occurring before menopause, HRT exposure, and age at menopause are associated with gut dysbiosis and insulin homeostasis. We hypothesized that gynecologic surgery before menopause, lack of HRT, and menopause age at age <50 years are associated with dysglycemia, defined as fasting plasma glucose >100 mg/dL or post-prandial plasma glucose >140 mg/dL at 2 hours in an oral glucose tolerance test, and insulin-related traits including insulin sensitivity, beta-cell function as measured by the disposition index, and insulin clearance. We hypothesized further that these menopause-related factors are associated with markers of gut dysbiosis and microbiota-associated plasma metabolites. In post-hoc analyses, we explored whether alterations in metabolites related to menopause factors and dysbiosis mediate differences in insulin sensitivity.

Materials and Methods

Participants

Data were collected from the baseline sample (n=353) of the MILES Study²¹. Recruitment into MILES has been reported elsewhere²¹. Briefly, exclusion criteria at enrollment included recent antibiotic or gastric acid secretion-suppressing medication use, current or planned pregnancy or breastfeeding, diabetes mellitus diagnosis, on-going severe illness, current oral steroid use, history of bariatric surgery, and chronic diarrhea²¹. Participants who acquired any of these criteria prior to the baseline study visit were omitted from analysis (Supplemental Figure 1).

Soluble Fiber Intake and Physical Activity Assessments

Soluble fiber consumption in grams/day, a key substrate for intestinal microbial activity²³, over the past year was inferred from the Diet History Questionnaire II²¹. Physical activity in metabolic equivalent minutes/week was estimated by survey using the Multi-Ethnic Study of Atherosclerosis Typical Week Physical Activity Survey²¹.

Menopause History Assessment

Menopause status, age at hysterectomy and/or bilateral oophorectomy, menses status prior to either surgery, menopause age, and history of HRT were obtained by participant questionnaires completed at the study visit.

Dysglycemia and Insulin Homeostasis Traits

The oral glucose tolerance test (OGTT) in MILES was performed as previously described²¹. In short, plasma glucose, insulin, and c-peptide levels were measured in the fasting state and post-prandial at 30 minutes and 2 hours after administration of a 75 g oral glucose load. Prediabetes was defined as having either a fasting plasma glucose of 100–125 mg/dL or post-prandial plasma glucose of 140–199 mg/dL at the 2-hour timepoint. Diabetes was defined as having both a fasting plasma glucose ≥126 mg/dL and a post-prandial plasma glucose ≥200 mg/dL at the 2-hour timepoint. Dysglycemia was defined as having any of fasting plasma glucose >100 mg/dL or post-prandial plasma glucose >140 mg/dL at the 2-hour timepoint. Insulin sensitivity was assessed with the Matsuda index²⁴. Insulin secretion was inferred by the area under the curve formed by insulin divided by the area under the curve of glucose at 30 minutes²⁵. The disposition index was taken as the product of the insulin sensitivity index and insulin secretion at 30 minutes²⁵. Insulin clearance was estimated by the area under the curve of insulin divided by the area under the curve for c-peptide over 2 hours²⁶.

Fecal Microbiota Taxonomic Profiling

Fecal microbiota sampling, sequencing, and taxonomic classification in MILES have been previously described²⁷. Briefly, participants provided fecal samples using the OMNIgene GUT self-collection kit. Whole-metagenome shotgun sequencing at a depth sufficient for species-level resolution was performed using the HiSeqX paired-end (2×150 base pair) platform. Low-quality, human genome, and PhiX sequences were removed before taxonomic classification and estimated read count quantification using MetaPhlAn3²⁸ and the “-rel_ab_w_read_stats” option. High-quality reads that failed MetaPhlAn3 taxonomic classification were separately binned and retained in the final taxon count table to preserve sample read depth prior to each analysis unless otherwise stated (see Supplement).

Plasma Metabolite Quantification

Plasma was collected following overnight fasting. Untargeted metabolite levels and targeted SCFA levels (acetate, propionate, butyrate, 2-methyl-butyrate, isobutyrate, and hexanoate) were quantified by ultra-high performance liquid chromatography tandem mass spectrometry (Metabolon, Morrisville, North Carolina) as previously described^27,29.

Microbiota Feature Selection

Species count data and clinical-translational metadata were merged using Phyloseq³⁰. A prevalence threshold was applied to include species present in at least 15% of the cohort (see Supplement). Default parameters were used in each analytical step unless otherwise stated.

Beta Diversity

Variation in between-sample phylogenetic composition was assessed using the Phylogenetic Robust Aitchison Principal Components Analysis in Gemellli³¹ within Qiime2³² (see Supplement).

Microbiota Co-abundance Network

A network of co-abundant fecal microbial species was constructed using the Semi-Parametric Rank-based approach for Inference in Graphical model in NetCoMi³³ (see Supplement). Microbial modules, defined as subsets of densely positively co-abundant species, were inferred using the igraph fast-greedy clustering algorithm on a Weighted Gene Network Correlation Network Analysis topologic overlap dissimilarity matrix in NetCoMi. Module relative abundance scores were calculated as the first principal components axis³⁴ of module member counts after centered-log ratio transformation.

Statistical Analysis

All analyses were performed in cross-section using the baseline sample in MILES. The primary independent variables, gynecologic surgery before menopause, history of HRT, or menopause age <50 years, were each modeled as binary categorical variables. Data transformations were performed on continuous variables to control skew (see Supplement).

Covariate selection was informed by development of a directed acyclic graph (DAG) using DAGitty³⁵. When gynecologic surgery history served as the primary exposure, self-reported Black or White race was modeled as a covariate. When HRT history was the primary exposure, menopause age <50 years, surgery history, and race were modeled as covariates. When menopause age was the primary exposure, models were adjusted for surgery history and chronological age. In post-hoc analyses where a microbiota variable and dysglycemia or an insulin trait was modeled as an exposure and outcome, respectively, adjusted models included soluble fiber intake-grams and metabolic equivalent minutes/week to account for confounding in diet and physical activity. Adjusted-model-1 adjusted for the above DAG-informed covariates. Adjusted-model-2 included adjusted-model-1 covariates plus BMI at study visit given potential confounding from BMI and recognizing that BMI could also be a mediator within the causal pathway. Batch-normalized creatinine was added as a covariate when plasma metabolites were modeled as an outcome. Where possible, missing menopause status and missing menopause age responses were derived from alternative data elements. Otherwise missing data were multiple imputed (see Supplement).

When insulin homeostasis traits, dysglycemia, or SCFA served as the outcome, generalized additive models of location, scale, and shape (GAMLSS)³⁶ were used for generalized linear or logistic regression with the inverse Gaussian distribution family or binomial family, respectively. In GAMLSS, only the location parameter was modeled against covariates. Beta-diversity using Gemelli was modeled against covariates by distance-based redundancy analysis within vegan. Quantile regression was performed when microbial co-abundance module scores served as the outcome using quant.reg. Boosted generalized additive modeling in MaAslin2³⁷ (Gaussian distribution) was performed for outcomes of Box-Cox-transformed plasma metabolites. Individual species differential abundance was modeled using Analysis of Compositions of Microbiomes with Bias Correction-2 (ANCOM-BC2)³⁸ with sensitivity testing enabled. Tests of mediation were performed using mediation³⁹ (see Supplement).

Statistical significance was assessed by 95% confidence interval (CI), as P<0.050, or Q<0.100 with Q-values calculated via the False Discovery Rate of Benjamini and Hochberg⁴⁰. R-based statistical packages were used for all analyses (see Supplement).

Results

Of the 218 participants reporting female sex, 175 reported being post-menopause and were included in data analyses after imputation of missing data (Supplemental Figure 1). These participants had a median (IQR) age of 61 (11) years, with 61% non-Hispanic White, 94% medically insured, 91% having some or more college education, and 68% with no history of HRT (Table 1). Hysterectomy or bilateral oophorectomy occurring before menopause was reported in 29% of participants. Black females comprised 39% of the cohort and reported a history of pre-menopausal gynecologic surgery more frequently than non-Black females (57% vs. 31%, P=0.003, Table 1). HRT use was more commonly reported among participants with a history of pre-menopausal gynecologic surgery (43% vs. 27%, Table 1) without reaching statistical significance (P=0.065). Median HRT duration was longer in the surgery group (10 vs. 5 years, P=0.008, Table 1). Menopause age at <50 years was more common in the gynecologic surgery group compared to the no surgery group (80% vs. 34%, P<0.001able). Among the pre-menopausal surgery group, 98% underwent hysterectomy and 53% underwent bilateral oophorectomy. BMI (29 vs. 26 kg/m², P=0.011) was statistically significantly greater and soluble fiber grams intake per day (5.02 g/day vs. 6.78 g/day, P=0.008) lower in the surgery group (Table 1). Physical activity in metabolic equivalents per week (1,410 MET/week vs. 1,504 MET/week, P=0.427) was similar between groups (Table 1).

Table 1.

Cohort characteristics stratified by history of gynecologic surgery (hysterectomy and/or bilateral oophorectomy) prior to menopause.

	Total Median or N (IQR or %) N=175	Surgery Prior to Menopause Median or N (IQR or %) N=51	No Surgery Prior to Menopause Median or N (IQR or %) N=124	P^†
Demographics

Age at Visit, years	61 (11)	58 (14)	61 (9)	0.191
African American	68 (39)	29 (57)	39 (31)	0.003
Body Mass Index, kg/m²	27.438 (9)	29 (11)	26 (9)	0.011
Any Medical Insurance	165 (94)	47 (92)	118 (95)	0.675

Education

Before High School	1 (0.6)	0 (0)	1 (0.8)	0.814
High School	14 (8)	4 (7.8)	10 (8.1)
Some College	58 (33.1)	20 (39.2)	38 (30.6)
Bachelors Degree	57 (32.6)	15 (29.4)	42 (33.9)
Graduate or Professional Degree	45 (25.7)	12 (23.5)	33 (26.6)

Income

<12,000	10 (5.7)	1 (2)	9 (7.3)	0.713
12–24,999	12 (6.9)	5 (9.8)	7 (5.6)
25–39,999	28 (16)	11 (21.6)	17 (13.7)
40–49,999	18 (10.3)	6 (11.8)	12 (9.7)
50–74,999	33 (18.9)	11 (21.6)	22 (17.7)
75–99,999	21 (12)	5 (9.8)	16 (12.9)
100–124,999	11 (6.3)	3 (5.9)	8 (6.5)
125–149,999	7 (4)	2 (3.9)	5 (4)
>150,000	14 (8)	3 (5.9)	11 (8.9)
Withheld	21 (12)	4 (7.8)	17 (13.7)

Lifestyle Factors

Soluble Fiber, g/day	6.24 (4.66)	5.02 (3.79)	6.78 (4.79)	0.008
Metabolic Equivalents, MET/week	1470 (1838)	1410 (2168)	1504 (1702)	0.427

Menopause Traits

History of HRT	56 (32)	22 (43)	34 (27)	0.065
Duration on HRT, years	7 (8)	10 (10)	5 (8)	0.008
Menopause Age <50 years	83 (47)	41 (80)	42 (34)	<.001

Glycemic Traits

Pre-diabetes^‡	71 (41)	25 (49)	46 (37)	0.197
Dysglycemia^§	74 (42)	27 (53)	47 (38)	0.097
Insulin Sensitivity Index (ISI)	4.451 (4.369)	3.819 (2.313)	4.808 (4.796)	0.056
Insulin Disposition Index (ISI x AUC-Ins₃₀/AUC-Glu₃₀)	1.560 (1.213)	1.604 (1.535)	1.555 (0.984)	0.621
Insulin Clearance Index (AUC-Cpep₁₂₀/AUC-Ins₁₂₀)	0.108 (0.052)	0.108 (0.045)	0.108 (0.052)	0.226

Open in a new tab

^†

P-values were calculated by Pearson’s χ² or Mann-Whitney test where appropriate.

^‡

Defined as having either a fasting plasma glucose of 100–125 mg/dL or post-prandial plasma glucose of 140–199 mg/dL at 2 hrs in an oral glucose tolerance test.

^§

Defined as fasting plasma glucose >100 mg/dL or post-prandial plasma glucose >140 mg/dL at 2 hrs in an oral glucose tolerance test. Abbreviations: hormone replacement therapy (HRT)

Menopause factors and insulin homeostasis traits

Dysglycemia was not statistically significantly associated with any menopause factor (Table 2). In contrast, insulin sensitivity index (ISI) scores were significantly lower in those reporting a history of pre-menopausal gynecologic surgery compared to no surgery in the unadjusted (β=−0.261, CI=[−0.474,−0.048]) model only (Table 2). Use of HRT was associated with lower ISI in adjusted-model-2 only (β=−0.275, CI=[−0.444,−0.107]). Menopause age at <50 years was associated with decreased ISI in the unadjusted model (β=−0.287, CI=[−0.489,−0.086]) and adjusted-model-1 (β=−0.232, CI=[−0.450,−0.014]; Table 2). HRT use was associated with a reduced disposition index in adjusted-model-2 (β=−0.225, CI=[−0.419,−0.032]; Table 2) only. Lastly, menopause age at <50 years was associated with reduced insulin clearance in the unadjusted model (β=−0.117, CI=[−0.228,−0.007]), only (Table 2).

Table 2. Menopause factors are associated with insulin homeostasis traits.

Bold indicates statistical significance was met based on 95% confidence interval boundaries and a null beta-coefficient of zero. Hysterectomy and/or bilateral oophorectomy before menopause (gynecologic surgery), history of hormone replacement therapy (HRT) use, and menopause age <50 years were compared to the no surgery before menopause, no history of HRT, and menopause age ≥50 years referent groups, respectively. Adjusted-model-1 covariates are as follows: gynecologic surgery adjusted for self-reported race; HRT use adjusted for self-reported race, gynecologic surgery history, and menopause age <50 years; and menopause age <50 years adjusted for age at study visit and gynecologic surgery history. Adjusted-model-2 covariates include adjusted-model-1 covariates with additional adjustment for body mass index.

	Unadjusted			Adjusted-model-1			Adjusted-model-2
	2.5% CI	β	97.5% CI	2.5% CI	β	97.5% CI	2.5% CI	β	97.5% CI
Dysglycemia ^†

Gynecologic surgery	−0.047	0.611	1.270	−0.190	0.489	1.168	−0.260	0.431	1.123
HRT use	−0.608	0.034	0.677	−0.605	0.079	0.763	−0.463	0.249	0.960
Menopause age <50 yrs	−0.235	0.367	0.970	−0.405	0.285	0.974	−0.550	0.164	0.877

Insulin Sensitivity Index

Gynecologic surgery	−0.474	−0.261	−0.048	−0.198	0.020	0.237	−0.075	0.106	0.288
HRT use	−0.308	−0.089	0.129	−0.320	−0.123	0.074	−0.444	−0.275	−0.107
Menopause age <50 yrs	−0.489	−0.287	−0.086	−0.450	−0.232	−0.014	−0.261	−0.099	0.063

Insulin Disposition Index

Gynecologic surgery	−0.207	−0.006	0.195	−0.207	0.013	0.233	−0.167	0.044	0.256
HRT use	−0.325	−0.135	0.056	−0.349	−0.152	0.044	−0.419	−0.225	−0.032
Menopause age <50 yrs	−0.212	−0.028	0.155	−0.240	−0.041	0.157	−0.237	−0.050	0.138

Insulin Clearance Index

Gynecologic surgery	−0.205	−0.084	0.036	−0.103	0.007	0.116	−0.066	0.039	0.145
HRT use	−0.094	0.027	0.147	−0.136	−0.027	0.082	−0.181	−0.075	0.031
Menopause age <50 yrs	−0.228	−0.117	−0.007	−0.230	−0.103	0.024	−0.148	−0.034	0.080

Open in a new tab

^†

Defined as fasting plasma glucose >100 or post-prandial plasma glucose >140 mg/dL at 2 hrs during an oral glucose tolerance test. Beta regression coefficients are listed with 95% confidence interval bounds.

Fecal microbiota composition and pre-menopausal gynecologic surgery

Beta-diversity Analysis

Gut microbiota beta-diversity or the phylogenetic variation in microbial species composition between fecal samples was assessed using Gemelli, a taxonomically-informed, abundance-sensitive, difference statistic robust to the influence of compositionality and unequal microbiota sampling³¹. Pre-menopausal gynecologic surgery was associated with beta-diversity and clustered samples significantly from those without a history of surgery in the unadjusted model (F=5.842, P=0.003), adjusted-model-1 (F=6.247, P=0.002), and adjusted-model-2 (F=6.120, P=0.002; Figure 1). HRT use was also significantly associated with sample beta-diversity in the unadjusted model (F=3.279, P=0.037) only (Figure 1). In contrast, menopause age <50 years failed to significantly cluster fecal samples (Figure 1).

Non-metric multidimensional scaling (NMDS) was performed to summarize the between-participant differences in fecal microbiota phylogenetic composition based on the Gemelli metric comparing A) hysterectomy and/or bilateral oophorectomy before menopause to no surgery, B) history of hormone replacement therapy (HRT) use to no use, and C) menopause age <50 years to ≥50 years. Each point represents a single participant and are colorized in black or gold according to their respective menopause factor class. Between-point spacing is determined by each participant’s similarity in fecal phylogenetic composition. Standard-error 95% confidence ellipses about the centroid for each population represented by black or gold lines are plotted. P-values adjusted for covariates in adjusted-model-2 were calculated by permutation test using Adonis2 in *vegan* in the R environment. Adjusted-model-2 covariates are as follows: gynecologic surgery adjusted for self-reported race and body mass index; HRT use adjusted for self-reported race, gynecologic surgery history, menopause age <50 years, and body mass index; and menopause age <50 years adjusted for age at study visit, gynecologic surgery history, and body mass index.

Co-abundance Network Analysis

Differences in microbiota composition were investigated by grouping microbial species according to their observed co-abundance patterns. A total of 11 groups referred to as modules of densely, positively co-abundant microbial species were observed (Figure 2, Supplemental Figure 2). The two modules largest in species membership were the green (39 species) and magenta (36 species) modules (Figure 2).

A) A network diagram showing inter-species co-abundance relationships as inferred by the Semi-Parametric Rank-based approach for INference in Graphical model (SPRING). Nodes represent individual species. Only positive co-abundance relationships are plotted and are represented by grey edges connecting nodes. Only the within-module edges have been plotted to simplify visualization, and the fully connected network may be found in Supplemental Figure 2. Species with dense co-abundance relationships were classified into modules using the Weighted Gene Correlation Network Analysis topologic overlap dissimilarity measure and greedy clustering in NetCoMi. B-C) Module abundance scores were obtained by taking the first principal component of centered-log ratio transformed species counts. The rank-percentile of the B) green and C) magenta module abundance scores are plotted as a function of gynecologic surgery history before menopause. The beta coefficient and 95% confidence interval (CI) represent the effect of having a history of surgery compared to no surgery on the median of the module abundance scores in a quantile regression model adjusting for the covariates in adjusted-model-2. Adjusted-model-2 covariates are as follows: gynecologic surgery adjusted for self-reported race and body mass index; HRT use adjusted for self-reported race, gynecologic surgery history, menopause age <50 years, and body mass index; and menopause age <50 years adjusted for age at study visit, gynecologic surgery history, and body mass index.

In the surgery group, the green module relative abundance score was significantly statistically decreased in the unadjusted model (β=−1.299, CI=[−2.313,−0.159]), adjusted-model-1 (β=−1.300, CI=[−2.515,−0.162]), and the adjusted-model-2 (β=−1.542, CI=[−2.392,−0.250]; Figure 2, Supplemental Table 1). Conversely, the surgery group was positively associated with the magenta module scores in the unadjusted model (β=2.349, CI=[0.728,4.162]), adjusted-model-1 (β=2.572, CI=[0.751,4.111]), and adjusted-model-2 (β=2.750, CI=[0.978,4.013]; Figure 2, Supplemental Table 1). In post-hoc modeling, dysglycemia was positively associated with the magenta module in adjusted-model-2 (β=0.008, CI=[0.001,0.016]) whereas the ISI trended towards a negative association with the magenta module in adjusted-model-2 (β=−0.001, CI=[−0.003,<0.001] without meeting statistical significance, data not shown).

Menopause age at <50 years was associated with decreased green module relative abundance scores (β=−1.280, CI=[−1.980,−0.346]) and increased cyan module scores (β=0.649, CI=[0.003,1.262]; Supplemental Table 1) in the unadjusted model. Menopause age was positively associated with pink module scores in the unadjusted model (β=1.060, CI=[0.151,1.692]) and adjusted-model-1 (β=0.867, CI=[0.015,1.563], Supplemental Table 1). . Additional associations were observed in tests of individual species differential abundance via ANCOM-BC2 (Supplemental Figure 3 & Supplemental Table 2).

Pre-menopausal gynecologic surgery and fecal microbiota-associated plasma metabolites

Plasma metabolomic profiling supplemented by targeted measurement of short chain fatty-acids was performed to characterize the variation in circulating metabolites as a function of menopause factors and gut dysbiosis. Acetate was detected in all samples and was significantly lower in participants reporting a history of gynecologic surgery before menopause in the unadjusted (β=−0.186, CI=[−0.339,−0.032]) and adjusted-model-1 (β=−0.154, CI=[−0.307,−0.001]; Supplemental Table 3). Similarly, acetate was decreased in participants with a menopause age <50 years in the unadjusted (β=−0.186, CI=[−0.334,−0.037]) model only (Supplemental Table 3).

Isobutyrate was detected in 46% (n=80) of participants. A history of HRT was significantly positively associated with the odds of detecting isobutyrate in plasma in the unadjusted (β=0.788, CI=[0.140,1.435]), adjusted-model-1 (β=0.834, CI=[0.173,1.495]), and adjusted-model-2 (β=1.001, CI=[0.299,1.702]; Supplemental Table 3). No other menopause trait was associated with SCFA levels (Supplemental Table 3).

A total of 1,179 and 973 plasma metabolites before and after filtering for >50% cohort prevalence were measured independently of SCFA. The metabolites that met the prevalence criterion were tested for association with menopause-related factors after further filtering for metabolites that were also associated with any one microbial module using MaAslin2. The green module exhibited the largest total number of associations with 15 metabolites (Figure 3C). In the unadjusted and adjusted-model-1, pre-menopausal gynecologic surgery was associated with the levels of 23 and 18 metabolites, respectively, and decreased to 13 in adjusted-model-2 (Figure 3, Supplemental Table 4). Members of the plasmalogen family of glycerophospholipids were negatively associated with a history of gynecologic surgery in each model (Figure 3A, Supplemental Table 4). In contrast, a history of HRT was positively associated with two lysophospholipid metabolites and pantothenate (Figure 3B, Supplemental Table 4). Menopause age <50 years was associated with members of the acetylated or modified peptide class (Figure 3B, Supplemental Table 4).

Metabolites were first filtered for cohort prevalence of at least 50% and association with at least one fecal microbial module abundance score prior to testing for association with menopause factors in MaAslin2. Hysterectomy and/or bilateral oophorectomy before menopause (gynecologic surgery), history of hormone replacement therapy (HRT) use, and menopause age <50 years were compared to the no surgery before menopause, no history of HRT, and menopause age ≥50 years referent groups, respectively. A-C) Beta-coefficients from MaAslin2 are plotted of metabolites significantly associated with a history of A) gynecologic surgery before menopause, B) HRT use, and C) menopause age <50 years after covariate adjustment in adjusted-model-2 and p-value adjustment for multiple comparisons (Q<0.100). Adjusted-model-2 covariates are as follows: gynecologic surgery adjusted for self-reported race and body mass index and HRT use adjusted for self-reported race, gynecologic surgery history, menopause age <50 years, and body mass index. Batch-normalized creatinine was also added as a covariate to account for relative differences in renal function. MaAsLin2 beta coefficients represent the effect of the presence of each menopause factor compared to its absence on the ratio of each metabolite’s serum concentration to its batch median. D) A network outlining associations between microbiota co-abundance modules and metabolites is plotted. Node colors correspond with module colors. Black nodes represent metabolites. Gold and blue edges indicate positive and negative associations, respectively, between modules and metabolites.

Plasmalogen lipid metabolites and insulin sensitivity

Post-hoc structural equation modeling was performed to assess for potential mediators of insulin sensitivity as a function of menopause traits in adjusted-model-2. The metabolites in Figure 3 were tested as mediators. Two plasmalogens, 1-stearyl-2-linoleoyl-GPC (O-18:0/18:2) and 1-(1-enyl-stearoyl)-2-linoleoyl-GPC (P-18:0/18:2), were found to be significant mediators of an inverse association between pre-menopausal gynecologic surgery and the ISI (Average Causal Mediation Effect [ACME]=−0.085, P=0.005 & ACME=−0.090, P=0.004, respectively, data not shown). The bile acid metabolite, taurolithocholate 3-sulfate, and additional metabolites lactosyl-N-nervonoyl-sphingosine (d18:1/24:1) and cinnamoylglycine were also found to be significant mediators of the inverse association between pre-menopausal gynecologic surgery and the ISI (ACME=−0.056, P=0.025; ACME=−0.055, P=0.035; ACME=−0.051, P=0.044, respectively, data not shown). No other metabolite was found to be a statistically significant mediator.

Discussion

In this analysis of post-menopausal women in MILES, we observed that menopause occurring before age 50 years was associated with reduced insulin sensitivity based on the ISI, supporting a model whereby risk for incident diabetes tends to increase in menopause when menopause occurs at an earlier age⁴¹. Notably, this reduction in insulin sensitivity did not amount to an increase in odds for being dysglycemic in this cohort of non-diabetic women. Interestingly, a history of prior HRT use was associated with reduced insulin sensitivity and beta-cell function represented by the disposition index after adjustment for covariates including BMI. Prior studies have shown a reduction in diabetes risk with HRT use⁴², and our contrasting results may reflect a pattern of reverse causality whereby the most symptomatic women may have been more likely to have received HRT. After covariate adjustment, pre-menopausal hysterectomy and/or bilateral oophorectomy was not associated with insulin sensitivity, suggesting that a lower menopause age itself, often occurring in women undergoing pre-menopausal gynecologic surgery, is more closely associated with a greater risk for insulin resistance.

We investigated whether gut dysbiosis was associated with menopause-associated factors in a manner that might contribute to risk for diabetes. We did not find a consistent association between a menopause age of <50 years and gut dysbiosis as assessed by variation in microbial beta-diversity, composition, or microbiota-associated plasma metabolite levels.

However, we did observe an association between a history of gynecologic surgery before menopause and markers of dysbiosis. Surgery was associated with fecal beta-diversity suggesting that differences in gut microbial phylogenetic composition distinguish this population from females who did not undergo surgery. A microbial co-abundance analysis revealed that two large modules of co-abundant microbial species account for these differences. Of the two modules, the relative abundance of the magenta module was greater in the surgery group and was dominated by a densely co-abundant cohort of Clostridium species such as C. symbiosum, C. bolteae, and C. citroniae in addition to Ruminococcus gnavus and Blautia wexlerae. Notably, the co-abundance of C. bolteae, C. citroniae, and R. gnavus in human feces has been previously reported as a risk factor for incident T2DM in a longitudinal Finnish study with a 15-year median follow up period by Ruuskanen et. al.⁴³. Moreover, metabolic byproducts of R. gnavus have been shown in pre-clinical models to directly impair insulin sensitivity⁴⁴. In MILES, the presence of dysglycemia was associated with augmented magenta module relative abundance suggesting that this microbial population may be more broadly associated with impaired glucose tolerance in post-menopausal females. In contrast, B. wexlerae appears to attenuate risk for T2DM in pre-clinical models⁴⁵, and its greater relative abundance in the surgery group in MILES may counter-balance the negative impact of the magenta module on insulin sensitivity.

Conversely, the relative abundance of the green module tended to decrease in the gynecologic surgery group and consisted of a more phylogenetically diverse set of co-abundant species including Alistipes indistinctus, which was associated with protection against incident T2DM in Ruuskanen et. al consistent with its potential to promote insulin sensitivity in pre-clinical models⁴⁶. Akkermansia muciniphila within the green module has been reported extensively as protective against dysglycemia in pre-clinical models and, more recently, a clinical trial⁴⁷.

In MILES, gynecologic surgery was associated with altered levels of plasma metabolites with reductions observed in plasmalogen glycerophospholipids. An analogous reduction in plasmalogens has been reported in post-menopausal women with HIV⁴⁸. Plasmalogens are host-derived plasma membrane-acting, peroxisome-sequestered lipids necessary for cellular signaling and function, and reductions in circulating levels have been shown to accompany insulin resistance⁴⁹ and incident T2DM^50,51. Gut microbiota, including Clostridia⁵², Bifidobacteria⁵³, and Streptococci⁵¹ also synthesize plasmalogens, which are thought to reduce oxidative stress in bacteria⁵⁴ and promote colonization^54,55. We speculate that a capacity to buffer against host plasmalogen deficiency could constitute a niche for Clostridia to expand in the gut and suppress the abundance of genera that might otherwise promote insulin sensitivity (i.e. Alistipes). Additional studies are needed to test this hypothesis and assess the therapeutic potential of plasmalogens in T2DM prevention⁵¹. Together, these data suggest that changes in the gut microbiota of non-diabetic women who undergo pre-menopausal gynecologic surgery resembles the dysbiosis observed in persons at risk for incident T2DM.

The findings herein are limited by the cross-sectional study design, which precludes drawing conclusions of causality. The dependence on recall for ascertaining menopause histories underlies a potential for misclassification or recall bias. Additionally, oophorectomy and hysterectomy histories were consolidated into a single group, and future, larger studies would benefit by stratifying these to clarify their differential contributions. Differences in plasma metabolites in MILES, including plasmalogens, may be driven by alterations in host or microbe synthesis, utilization, and excretion and may be better interrogated by measuring metabolic kinetics endovascularly and in the colon. Future studies utilizing a larger sample size and incorporating external validation datasets are warranted given the potential for spurious findings in testing multiple hypotheses.

In summary, we show that menopause age <50 years is associated with insulin resistance whereas pre-menopausal hysterectomy and/or bilateral oophorectomy is more closely associated with gut dysbiosis. Our findings suggest a model by which gynecologic surgery before menopause could contribute to impaired insulin homeostasis as manifested in alterations in gut microbial species and circulating plasmalogens that support host metabolism. Future trials could evaluate these targets for their potential to mitigate risk for insulin resistance in women who undergo pre-menopausal gynecological surgery and/or early menopause.

Supplementary Material

Supinfo1

NIHMS2093530-supplement-Supinfo1.docx^{(8.2KB, docx)}

Supinfo2

Supplemental Figure 1 Diagram of the sample selection from the MILES baseline cohort as described in the materials and methods. Only three participants were excluded in this analysis after enrollment as one participant met two of the listed exclusion criteria.

NIHMS2093530-supplement-Supinfo2.pdf^{(132.4KB, pdf)}

Supinfo5

Supplemental Table 1 Menopause factors are associated with co-abundance module relative abundance. Module color names correspond to node color in Figure 2 and Supplemental Figure 2. Bold indicates statistical significance was met based on 95% confidence interval boundaries and a null beta-coefficient of zero. Hysterectomy and/or bilateral oophorectomy before menopause (gynecologic surgery), history of hormone replacement therapy (HRT) use, and menopause age <50 years were compared to the no surgery before menopause, no history of HRT, and menopause age ≥50 years referent groups, respectively. Adjusted-model-1 covariates are as follows: gynecologic surgery adjusted for self-reported race; HRT use adjusted for self-reported race, gynecologic surgery history, and menopause age <50 years; and menopause age <50 years adjusted for age at study visit and gynecologic surgery history. Adjusted-model-2 covariates include adjusted-model-1 covariates with additional adjustment for body mass index. Beta coefficients represent the change in median module abundance score associated with each menopause factor.

NIHMS2093530-supplement-Supinfo5.xlsx^{(9.8KB, xlsx)}

Supinfo4

Supplemental Figure 3 Fecal microbial species are differentially abundant according to menopause factors. Analysis of Compositions of Microbiomes with Bias Correction-2 (ANCOM-BC2) was used to detect statistically significant differences in microbial relative abundances among menopause factors. Hysterectomy and/or bilateral oophorectomy before menopause (gynecologic surgery), history of hormone replacement therapy (HRT) use, and menopause age <50 years were compared to the no surgery before menopause, no history of HRT, and menopause age ≥50 years referent groups, respectively. Adjusted-model-1 covariates are as follows: gynecologic surgery adjusted for self-reported race; HRT use adjusted for self-reported race, gynecologic surgery history, and menopause age <50 years; and menopause age <50 years adjusted for age at study visit and gynecologic surgery history. Adjusted-model-2 covariates include adjusted-model-1 covariates with additional adjustment for body mass index. Abbreviations: NS-not significant; Log2-FC-log base 2 fold change was greater than 1; Q-value-FDR-adjusted p-value was <0.100; Log2-FC and q-value-both criteria were met; Ntax = number of species passing sensitivity testing; Nsamp = number of samples in the analysis.

NIHMS2093530-supplement-Supinfo4.pdf^{(635.1KB, pdf)}

Supinfo6

Supplemental Table 3 Menopause factors are associated with plasma short chain fatty acid concentrations. Beta regression coefficients are listed with 95% confidence interval bounds and represent the log-odds ratio for the detection of each short chain fatty-acid excluding acetate, which was modeled as a continuous variable. Bold indicates statistical significance was met based on 95% confidence interval boundaries and a null beta-coefficient of zero. Hysterectomy and/or bilateral oophorectomy before menopause (gynecologic surgery), history of hormone replacement therapy (HRT) use, and menopause age <50 years were compared to the no surgery before menopause, no history of HRT, and menopause age ≥50 years referent groups, respectively. Adjusted-model-1 covariates are as follows: gynecologic surgery adjusted for self-reported race; HRT use adjusted for self-reported race, gynecologic surgery history, and menopause age <50 years; and menopause age <50 years adjusted for age at study visit and gynecologic surgery history. Adjusted-model-2 covariates include adjusted-model-1 covariates with additional adjustment for body mass index.

NIHMS2093530-supplement-Supinfo6.xlsx^{(8.3KB, xlsx)}

Supinfo7

Supplemental Table 4 Menopause factors are associated with plasma metabolite levels. Expanded table of tests of association between menopause factors and plasma metabolites that were prevalent in at least 50% of the cohort and associated with at least one microbial module. Only statistically significant (Q<0.100) results are shown. Beta coefficients, standard errors, and p-values were generated from MaAslin2. Hysterectomy and/or bilateral oophorectomy before menopause (gynecologic surgery), history of hormone replacement therapy (HRT) use, and menopause age <50 years were compared to the no surgery before menopause, no history of HRT, and menopause age ≥50 years referent groups, respectively. Adjusted-model-1 covariates are as follows: gynecologic surgery adjusted for self-reported race; HRT use adjusted for self-reported race, gynecologic surgery history, and menopause age <50 years; and menopause age <50 years adjusted for age at study visit and gynecologic surgery history. Adjusted-model-2 covariates include adjusted-model-1 covariates with additional adjustment for body mass index. Batch-normalized creatinine was also added as a covariate to both models to account for relative differences in renal function.

NIHMS2093530-supplement-Supinfo7.xlsx^{(12KB, xlsx)}

Supinfo3

Supplemental Figure 2 Complete fecal microbial co-abundance network with full length species names. Network edges are plotted in a Fruchterman-Reingold projection.

NIHMS2093530-supplement-Supinfo3.pdf^{(2.6MB, pdf)}

Supinfo8

Supplemental Table 2 Menopause factors are associated with fecal microbial species relative abundance based on ANCOM-BC2. Tests of association were performed using Analysis of Compositions of Microbiomes with Bias Correction-2 (ANCOM-BC2) and correspond to data presented in Supplemental Figure 3. Hysterectomy and/or bilateral oophorectomy before menopause (gynecologic surgery), history of hormone replacement therapy (HRT) use, and menopause age <50 years were compared to the no surgery before menopause, no history of HRT, and menopause age ≥50 years referent groups, respectively. Adjusted-model-1 covariates are as follows: gynecologic surgery adjusted for self-reported race; HRT use adjusted for self-reported race, gynecologic surgery history, and menopause age <50 years; and menopause age <50 years adjusted for age at study visit and gynecologic surgery history. Adjusted-model-2 covariates include adjusted-model-1 covariates with additional adjustment for body mass index.

NIHMS2093530-supplement-Supinfo8.xlsx^{(273.4KB, xlsx)}

Acknowledgments

Funding was provided by the National Institute of Diabetes and Digestive and Kidney Diseases (R01-DK109588, P30-DK063491) and the National Center for Advancing Translational Sciences (UL1-TR001420, UL1-TR001881). ACW received support from the U.S. Department of Agriculture cooperative agreement 58-3092-5-001. MOG was supported by the Eris M. Field Chair in Diabetes Research.

References

1.Moorman PG, Myers ER, Schildkraut JM, Iversen ES, Wang F, Warren N. Effect of Hysterectomy With Ovarian Preservation on Ovarian Function. Obstet Gynecol 2011;118(6):1271–1279. doi: 10.1097/AOG.0b013e318236fd12 [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Trabuco EC, Moorman PG, Algeciras-Schimnich A, Weaver AL, Cliby WA. Association of Ovary-Sparing Hysterectomy With Ovarian Reserve. Obstet Gynecol 2016;127(5):819–827. doi: 10.1097/AOG.0000000000001398 [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Madueke-Laveaux OS, Elsharoud A, Al-Hendy A. What We Know about the Long-Term Risks of Hysterectomy for Benign Indication-A Systematic Review. J Clin Med 2021;10(22):5335. doi: 10.3390/jcm10225335 [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Luo J, Manson JE, Urrutia RP, Hendryx M, LeBlanc ES, Margolis KL. Risk of Diabetes After Hysterectomy With or Without Oophorectomy in Postmenopausal Women. Am J Epidemiol 2017;185(9):777–785. doi: 10.1093/aje/kwx023 [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Desai S, Singh RJ, Govil D, et al. Hysterectomy and women’s health in India: evidence from a nationally representative, cross-sectional survey of older women. Womens Midlife Health. 2023;9(1):1. doi: 10.1186/s40695-022-00084-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Chiang CH, Chen W, Tsai IJ, et al. Diabetes mellitus risk after hysterectomy: A population-based retrospective cohort study. Medicine (Baltimore). 2021;100(4):e24468. doi: 10.1097/MD.0000000000024468 [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Hassan H, Allen I, Sofianopoulou E, et al. Long-term outcomes of hysterectomy with bilateral salpingo-oophorectomy: a systematic review and meta-analysis. Am J Obstet Gynecol 2024;230(1):44–57. doi: 10.1016/j.ajog.2023.06.043 [DOI] [PubMed] [Google Scholar]
8.Anagnostis P, Christou K, Artzouchaltzi AM, et al. Early menopause and premature ovarian insufficiency are associated with increased risk of type 2 diabetes: a systematic review and meta-analysis. Eur J Endocrinol 2019;180(1):41–50. doi: 10.1530/EJE-18-0602 [DOI] [PubMed] [Google Scholar]
9.Weyer C, Bogardus C, Mott DM, Pratley RE. The natural history of insulin secretory dysfunction and insulin resistance in the pathogenesis of type 2 diabetes mellitus. J Clin Invest 1999;104(6):787–794. doi: 10.1172/JCI7231 [DOI] [PMC free article] [PubMed] [Google Scholar]
10.De Paoli M, Zakharia A, Werstuck GH. The Role of Estrogen in Insulin Resistance: A Review of Clinical and Preclinical Data. Am J Pathol 2021;191(9):1490–1498. doi: 10.1016/j.ajpath.2021.05.011 [DOI] [PubMed] [Google Scholar]
11.Barros RPA, Gustafsson JÅ. Estrogen receptors and the metabolic network. Cell Metab 2011;14(3):289–299. doi: 10.1016/j.cmet.2011.08.005 [DOI] [PubMed] [Google Scholar]
12.Abildgaard J, Tingstedt J, Zhao Y, et al. Increased systemic inflammation and altered distribution of T-cell subsets in postmenopausal women. PloS One. 2020;15(6):e0235174. doi: 10.1371/journal.pone.0235174 [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Shieh A, Epeldegui M, Karlamangla AS, Greendale GA. Gut permeability, inflammation, and bone density across the menopause transition. JCI Insight. 2020;5(2):e134092, 134092. doi: 10.1172/jci.insight.134092 [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Rohm TV, Meier DT, Olefsky JM, Donath MY. Inflammation in obesity, diabetes, and related disorders. Immunity. 2022;55(1):31–55. doi: 10.1016/j.immuni.2021.12.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Miller AP, Chen YF, Xing D, Feng W, Oparil S. Hormone replacement therapy and inflammation: interactions in cardiovascular disease. Hypertens Dallas Tex 1979. 2003;42(4):657–663. doi: 10.1161/01.HYP.0000085560.02979.0C [DOI] [PubMed] [Google Scholar]
16.de Lauzon-Guillain B, Fournier A, Fabre A, et al. Menopausal hormone therapy and new-onset diabetes in the French Etude Epidemiologique de Femmes de la Mutuelle Générale de l’Education Nationale (E3N) cohort. Diabetologia. 2009;52(10):2092–2100. doi: 10.1007/s00125-009-1456-y [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Manson JE, Chlebowski RT, Stefanick ML, et al. Menopausal hormone therapy and health outcomes during the intervention and extended poststopping phases of the Women’s Health Initiative randomized trials. JAMA 2013;310(13):1353–1368. doi: 10.1001/jama.2013.278040 [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Marjoribanks J, Farquhar C, Roberts H, Lethaby A, Lee J. Long-term hormone therapy for perimenopausal and postmenopausal women. Cochrane Database Syst Rev 2017;1(1):CD004143. doi: 10.1002/14651858.CD004143.pub5 [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Suzuki T Regulation of intestinal epithelial permeability by tight junctions. Cell Mol Life Sci CMLS 2013;70(4):631–659. doi: 10.1007/s00018-012-1070-x [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Yang M, Wen S, Zhang J, Peng J, Shen X, Xu L. Systematic Review and Meta-analysis: Changes of Gut Microbiota before and after Menopause. Dis Markers. 2022;2022:3767373. doi: 10.1155/2022/3767373 [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Jensen ET, Bertoni AG, Crago OL, et al. Rationale, design and baseline characteristics of the Microbiome and Insulin Longitudinal Evaluation Study (MILES). Diabetes Obes Metab 2020;22(11):1976–1984. doi: 10.1111/dom.14145 [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Liu L, Fu Q, Li T, et al. Gut microbiota and butyrate contribute to nonalcoholic fatty liver disease in premenopause due to estrogen deficiency. PloS One. 2022;17(2):e0262855. doi: 10.1371/journal.pone.0262855 [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Guan ZW, Yu EZ, Feng Q. Soluble Dietary Fiber, One of the Most Important Nutrients for the Gut Microbiota. Mol Basel Switz 2021;26(22):6802. doi: 10.3390/molecules26226802 [DOI] [PMC free article] [PubMed] [Google Scholar]
24.DeFronzo RA, Matsuda M. Reduced time points to calculate the composite index. Diabetes Care. 2010;33(7):e93. doi: 10.2337/dc10-0646 [DOI] [PubMed] [Google Scholar]
25.Stancáková A, Javorský M, Kuulasmaa T, Haffner SM, Kuusisto J, Laakso M. Changes in insulin sensitivity and insulin release in relation to glycemia and glucose tolerance in 6,414 Finnish men. Diabetes. 2009;58(5):1212–1221. doi: 10.2337/db08-1607 [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Semnani-Azad Z, Johnston LW, Lee C, et al. Determinants of longitudinal change in insulin clearance: the Prospective Metabolism and Islet Cell Evaluation cohort. BMJ Open Diabetes Res Care. 2019;7(1):e000825. doi: 10.1136/bmjdrc-2019-000825 [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Cui J, Ramesh G, Wu M, et al. Butyrate-Producing Bacteria and Insulin Homeostasis: The Microbiome and Insulin Longitudinal Evaluation Study (MILES). Diabetes. 2022;71(11):2438–2446. doi: 10.2337/db22-0168 [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Beghini F, McIver LJ, Blanco-Míguez A, et al. Integrating taxonomic, functional, and strain-level profiling of diverse microbial communities with bioBakery 3. eLife. 2021;10:e65088. doi: 10.7554/eLife.65088 [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Li-Gao R, Grubbs K, Bertoni AG, et al. The Roles of Gut Microbiome and Plasma Metabolites in the Associations between ABO Blood Groups and Insulin Homeostasis: The Microbiome and Insulin Longitudinal Evaluation Study (MILES). Metabolites. 2022;12(9):787. doi: 10.3390/metabo12090787 [DOI] [PMC free article] [PubMed] [Google Scholar]
30.McMurdie PJ, Holmes S. Phyloseq: a bioconductor package for handling and analysis of high-throughput phylogenetic sequence data. Pac Symp Biocomput Pac Symp Biocomput Published online 2012:235–246. [PMC free article] [PubMed] [Google Scholar]
31.Martino C, Shenhav L, Marotz CA, et al. Context-aware dimensionality reduction deconvolutes gut microbial community dynamics. Nat Biotechnol 2021;39(2):165–168. doi: 10.1038/s41587-020-0660-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Bolyen E, Rideout JR, Dillon MR, et al. Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat Biotechnol 2019;37(8):852–857. doi: 10.1038/s41587-019-0209-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Peschel S, Müller CL, von Mutius E, Boulesteix AL, Depner M. NetCoMi: network construction and comparison for microbiome data in R. Brief Bioinform 2021;22(4):bbaa290. doi: 10.1093/bib/bbaa290 [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Langfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics. 2008;9:559. doi: 10.1186/1471-2105-9-559 [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Textor J, van der Zander B, Gilthorpe MS, Liskiewicz M, Ellison GT. Robust causal inference using directed acyclic graphs: the R package “dagitty.” Int J Epidemiol 2016;45(6):1887–1894. doi: 10.1093/ije/dyw341 [DOI] [PubMed] [Google Scholar]
36.Rigby RA, Stasinopoulos DM. Generalized additive models for location, scale and shape. J R Stat Soc Ser C Appl Stat 2005;54(3):507–554. doi: 10.1111/j.1467-9876.2005.00510.x [DOI] [Google Scholar]
37.Morgan XC, Tickle TL, Sokol H, et al. Dysfunction of the intestinal microbiome in inflammatory bowel disease and treatment. Genome Biol 2012;13(9):R79. doi: 10.1186/gb-2012-13-9-r79 [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Lin H, Peddada SD. Multigroup analysis of compositions of microbiomes with covariate adjustments and repeated measures. Nat Methods. 2024;21(1):83–91. doi: 10.1038/s41592-023-02092-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Tingley D, Yamamoto T, Hirose K, Keele L, Imai K. mediation: R Package for Causal Mediation Analysis. J Stat Softw 2014;59:1–38. doi: 10.18637/jss.v059.i0526917999 [DOI] [Google Scholar]
40.Glickman ME, Rao SR, Schultz MR. False discovery rate control is a recommended alternative to Bonferroni-type adjustments in health studies. J Clin Epidemiol 2014;67(8):850–857. doi: 10.1016/j.jclinepi.2014.03.012 [DOI] [PubMed] [Google Scholar]
41.Roa-Díaz ZM, Wehrli F, Lambrinoudaki I, et al. Early menopause and cardiovascular risk factors: a cross-sectional and longitudinal study. Menopause N Y N 2023;30(6):599–606. doi: 10.1097/GME.0000000000002184 [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Mauvais-Jarvis F, Manson JE, Stevenson JC, Fonseca VA. Menopausal Hormone Therapy and Type 2 Diabetes Prevention: Evidence, Mechanisms, and Clinical Implications. Endocr Rev 2017;38(3):173–188. doi: 10.1210/er.2016-1146 [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Ruuskanen MO, Erawijantari PP, Havulinna AS, et al. Gut Microbiome Composition Is Predictive of Incident Type 2 Diabetes in a Population Cohort of 5,572 Finnish Adults. Diabetes Care. 2022;45(4):811–818. doi: 10.2337/dc21-2358 [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Zhai L, Xiao H, Lin C, et al. Gut microbiota-derived tryptamine and phenethylamine impair insulin sensitivity in metabolic syndrome and irritable bowel syndrome. Nat Commun 2023;14(1):4986. doi: 10.1038/s41467-023-40552-y [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Hosomi K, Saito M, Park J, et al. Oral administration of Blautia wexlerae ameliorates obesity and type 2 diabetes via metabolic remodeling of the gut microbiota. Nat Commun 2022;13(1):4477. doi: 10.1038/s41467-022-32015-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Takeuchi T, Kubota T, Nakanishi Y, et al. Gut microbial carbohydrate metabolism contributes to insulin resistance. Nature. 2023;621(7978):389–395. doi: 10.1038/s41586-023-06466-x [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Depommier C, Everard A, Druart C, et al. Supplementation with Akkermansia muciniphila in overweight and obese human volunteers: a proof-of-concept exploratory study. Nat Med 2019;25(7):1096–1103. doi: 10.1038/s41591-019-0495-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Wang Y, Sharma A, Weber KM, et al. The menopause-related gut microbiome: associations with metabolomics, inflammatory protein markers, and cardiometabolic health in women with HIV. Menopause N Y N 2024;31(1):52–64. doi: 10.1097/GME.0000000000002287 [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Palmer ND, Okut H, Hsu FC, et al. Metabolomics Identifies Distinctive Metabolite Signatures for Measures of Glucose Homeostasis: The Insulin Resistance Atherosclerosis Family Study (IRAS-FS). J Clin Endocrinol Metab 2018;103(5):1877–1888. doi: 10.1210/jc.2017-02203 [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Razquin C, Toledo E, Clish CB, et al. Plasma Lipidomic Profiling and Risk of Type 2 Diabetes in the PREDIMED Trial. Diabetes Care. 2018;41(12):2617–2624. doi: 10.2337/dc18-0840 [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Beyene HB, Huynh K, Wang T, et al. Development and validation of a plasmalogen score as an independent modifiable marker of metabolic health: population based observational studies and a placebo-controlled cross-over study. EBioMedicine. 2024;105:105187. doi: 10.1016/j.ebiom.2024.105187 [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Jackson DR, Cassilly CD, Plichta DR, et al. Plasmalogen Biosynthesis by Anaerobic Bacteria: Identification of a Two-Gene Operon Responsible for Plasmalogen Production in Clostridium perfringens. ACS Chem Biol 2021;16(1):6–13. doi: 10.1021/acschembio.0c00673 [DOI] [PMC free article] [PubMed] [Google Scholar]
53.Mawatari S, Sasuga Y, Morisaki T, Okubo M, Emura T, Fujino T. Identification of plasmalogens in Bifidobacterium longum, but not in Bifidobacterium animalis. Sci Rep 2020;10(1):427. doi: 10.1038/s41598-019-57309-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
54.Joyce LR, Guan Z, Palmer KL. Streptococcus pneumoniae, S. pyogenes and S. agalactiae membrane phospholipid remodelling in response to human serum. Microbiol Read Engl 2021;167(5):001048. doi: 10.1099/mic.0.001048 [DOI] [PMC free article] [PubMed] [Google Scholar]
55.Mu R, Momeni S, Krieger M, et al. Plasmalogen, a glycerophospholipid crucial for Streptococcus mutans acid tolerance and colonization. Appl Environ Microbiol 2024;90(4):e0150023. doi: 10.1128/aem.01500-23 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supinfo1

NIHMS2093530-supplement-Supinfo1.docx^{(8.2KB, docx)}

Supinfo2

NIHMS2093530-supplement-Supinfo2.pdf^{(132.4KB, pdf)}

Supinfo5

NIHMS2093530-supplement-Supinfo5.xlsx^{(9.8KB, xlsx)}

Supinfo4

NIHMS2093530-supplement-Supinfo4.pdf^{(635.1KB, pdf)}

Supinfo6

NIHMS2093530-supplement-Supinfo6.xlsx^{(8.3KB, xlsx)}

Supinfo7

NIHMS2093530-supplement-Supinfo7.xlsx^{(12KB, xlsx)}

Supinfo3

Supplemental Figure 2 Complete fecal microbial co-abundance network with full length species names. Network edges are plotted in a Fruchterman-Reingold projection.

NIHMS2093530-supplement-Supinfo3.pdf^{(2.6MB, pdf)}

Supinfo8

NIHMS2093530-supplement-Supinfo8.xlsx^{(273.4KB, xlsx)}

[R1] 1.Moorman PG, Myers ER, Schildkraut JM, Iversen ES, Wang F, Warren N. Effect of Hysterectomy With Ovarian Preservation on Ovarian Function. Obstet Gynecol 2011;118(6):1271–1279. doi: 10.1097/AOG.0b013e318236fd12 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R2] 2.Trabuco EC, Moorman PG, Algeciras-Schimnich A, Weaver AL, Cliby WA. Association of Ovary-Sparing Hysterectomy With Ovarian Reserve. Obstet Gynecol 2016;127(5):819–827. doi: 10.1097/AOG.0000000000001398 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R3] 3.Madueke-Laveaux OS, Elsharoud A, Al-Hendy A. What We Know about the Long-Term Risks of Hysterectomy for Benign Indication-A Systematic Review. J Clin Med 2021;10(22):5335. doi: 10.3390/jcm10225335 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R4] 4.Luo J, Manson JE, Urrutia RP, Hendryx M, LeBlanc ES, Margolis KL. Risk of Diabetes After Hysterectomy With or Without Oophorectomy in Postmenopausal Women. Am J Epidemiol 2017;185(9):777–785. doi: 10.1093/aje/kwx023 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R5] 5.Desai S, Singh RJ, Govil D, et al. Hysterectomy and women’s health in India: evidence from a nationally representative, cross-sectional survey of older women. Womens Midlife Health. 2023;9(1):1. doi: 10.1186/s40695-022-00084-9 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R6] 6.Chiang CH, Chen W, Tsai IJ, et al. Diabetes mellitus risk after hysterectomy: A population-based retrospective cohort study. Medicine (Baltimore). 2021;100(4):e24468. doi: 10.1097/MD.0000000000024468 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R7] 7.Hassan H, Allen I, Sofianopoulou E, et al. Long-term outcomes of hysterectomy with bilateral salpingo-oophorectomy: a systematic review and meta-analysis. Am J Obstet Gynecol 2024;230(1):44–57. doi: 10.1016/j.ajog.2023.06.043 [DOI] [PubMed] [Google Scholar]

[R8] 8.Anagnostis P, Christou K, Artzouchaltzi AM, et al. Early menopause and premature ovarian insufficiency are associated with increased risk of type 2 diabetes: a systematic review and meta-analysis. Eur J Endocrinol 2019;180(1):41–50. doi: 10.1530/EJE-18-0602 [DOI] [PubMed] [Google Scholar]

[R9] 9.Weyer C, Bogardus C, Mott DM, Pratley RE. The natural history of insulin secretory dysfunction and insulin resistance in the pathogenesis of type 2 diabetes mellitus. J Clin Invest 1999;104(6):787–794. doi: 10.1172/JCI7231 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R10] 10.De Paoli M, Zakharia A, Werstuck GH. The Role of Estrogen in Insulin Resistance: A Review of Clinical and Preclinical Data. Am J Pathol 2021;191(9):1490–1498. doi: 10.1016/j.ajpath.2021.05.011 [DOI] [PubMed] [Google Scholar]

[R11] 11.Barros RPA, Gustafsson JÅ. Estrogen receptors and the metabolic network. Cell Metab 2011;14(3):289–299. doi: 10.1016/j.cmet.2011.08.005 [DOI] [PubMed] [Google Scholar]

[R12] 12.Abildgaard J, Tingstedt J, Zhao Y, et al. Increased systemic inflammation and altered distribution of T-cell subsets in postmenopausal women. PloS One. 2020;15(6):e0235174. doi: 10.1371/journal.pone.0235174 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R13] 13.Shieh A, Epeldegui M, Karlamangla AS, Greendale GA. Gut permeability, inflammation, and bone density across the menopause transition. JCI Insight. 2020;5(2):e134092, 134092. doi: 10.1172/jci.insight.134092 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R14] 14.Rohm TV, Meier DT, Olefsky JM, Donath MY. Inflammation in obesity, diabetes, and related disorders. Immunity. 2022;55(1):31–55. doi: 10.1016/j.immuni.2021.12.013 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R15] 15.Miller AP, Chen YF, Xing D, Feng W, Oparil S. Hormone replacement therapy and inflammation: interactions in cardiovascular disease. Hypertens Dallas Tex 1979. 2003;42(4):657–663. doi: 10.1161/01.HYP.0000085560.02979.0C [DOI] [PubMed] [Google Scholar]

[R16] 16.de Lauzon-Guillain B, Fournier A, Fabre A, et al. Menopausal hormone therapy and new-onset diabetes in the French Etude Epidemiologique de Femmes de la Mutuelle Générale de l’Education Nationale (E3N) cohort. Diabetologia. 2009;52(10):2092–2100. doi: 10.1007/s00125-009-1456-y [DOI] [PMC free article] [PubMed] [Google Scholar]

[R17] 17.Manson JE, Chlebowski RT, Stefanick ML, et al. Menopausal hormone therapy and health outcomes during the intervention and extended poststopping phases of the Women’s Health Initiative randomized trials. JAMA 2013;310(13):1353–1368. doi: 10.1001/jama.2013.278040 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R18] 18.Marjoribanks J, Farquhar C, Roberts H, Lethaby A, Lee J. Long-term hormone therapy for perimenopausal and postmenopausal women. Cochrane Database Syst Rev 2017;1(1):CD004143. doi: 10.1002/14651858.CD004143.pub5 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R19] 19.Suzuki T Regulation of intestinal epithelial permeability by tight junctions. Cell Mol Life Sci CMLS 2013;70(4):631–659. doi: 10.1007/s00018-012-1070-x [DOI] [PMC free article] [PubMed] [Google Scholar]

[R20] 20.Yang M, Wen S, Zhang J, Peng J, Shen X, Xu L. Systematic Review and Meta-analysis: Changes of Gut Microbiota before and after Menopause. Dis Markers. 2022;2022:3767373. doi: 10.1155/2022/3767373 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R21] 21.Jensen ET, Bertoni AG, Crago OL, et al. Rationale, design and baseline characteristics of the Microbiome and Insulin Longitudinal Evaluation Study (MILES). Diabetes Obes Metab 2020;22(11):1976–1984. doi: 10.1111/dom.14145 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R22] 22.Liu L, Fu Q, Li T, et al. Gut microbiota and butyrate contribute to nonalcoholic fatty liver disease in premenopause due to estrogen deficiency. PloS One. 2022;17(2):e0262855. doi: 10.1371/journal.pone.0262855 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R23] 23.Guan ZW, Yu EZ, Feng Q. Soluble Dietary Fiber, One of the Most Important Nutrients for the Gut Microbiota. Mol Basel Switz 2021;26(22):6802. doi: 10.3390/molecules26226802 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R24] 24.DeFronzo RA, Matsuda M. Reduced time points to calculate the composite index. Diabetes Care. 2010;33(7):e93. doi: 10.2337/dc10-0646 [DOI] [PubMed] [Google Scholar]

[R25] 25.Stancáková A, Javorský M, Kuulasmaa T, Haffner SM, Kuusisto J, Laakso M. Changes in insulin sensitivity and insulin release in relation to glycemia and glucose tolerance in 6,414 Finnish men. Diabetes. 2009;58(5):1212–1221. doi: 10.2337/db08-1607 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R26] 26.Semnani-Azad Z, Johnston LW, Lee C, et al. Determinants of longitudinal change in insulin clearance: the Prospective Metabolism and Islet Cell Evaluation cohort. BMJ Open Diabetes Res Care. 2019;7(1):e000825. doi: 10.1136/bmjdrc-2019-000825 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R27] 27.Cui J, Ramesh G, Wu M, et al. Butyrate-Producing Bacteria and Insulin Homeostasis: The Microbiome and Insulin Longitudinal Evaluation Study (MILES). Diabetes. 2022;71(11):2438–2446. doi: 10.2337/db22-0168 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R28] 28.Beghini F, McIver LJ, Blanco-Míguez A, et al. Integrating taxonomic, functional, and strain-level profiling of diverse microbial communities with bioBakery 3. eLife. 2021;10:e65088. doi: 10.7554/eLife.65088 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R29] 29.Li-Gao R, Grubbs K, Bertoni AG, et al. The Roles of Gut Microbiome and Plasma Metabolites in the Associations between ABO Blood Groups and Insulin Homeostasis: The Microbiome and Insulin Longitudinal Evaluation Study (MILES). Metabolites. 2022;12(9):787. doi: 10.3390/metabo12090787 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R30] 30.McMurdie PJ, Holmes S. Phyloseq: a bioconductor package for handling and analysis of high-throughput phylogenetic sequence data. Pac Symp Biocomput Pac Symp Biocomput Published online 2012:235–246. [PMC free article] [PubMed] [Google Scholar]

[R31] 31.Martino C, Shenhav L, Marotz CA, et al. Context-aware dimensionality reduction deconvolutes gut microbial community dynamics. Nat Biotechnol 2021;39(2):165–168. doi: 10.1038/s41587-020-0660-7 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R32] 32.Bolyen E, Rideout JR, Dillon MR, et al. Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat Biotechnol 2019;37(8):852–857. doi: 10.1038/s41587-019-0209-9 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R33] 33.Peschel S, Müller CL, von Mutius E, Boulesteix AL, Depner M. NetCoMi: network construction and comparison for microbiome data in R. Brief Bioinform 2021;22(4):bbaa290. doi: 10.1093/bib/bbaa290 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R34] 34.Langfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics. 2008;9:559. doi: 10.1186/1471-2105-9-559 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R35] 35.Textor J, van der Zander B, Gilthorpe MS, Liskiewicz M, Ellison GT. Robust causal inference using directed acyclic graphs: the R package “dagitty.” Int J Epidemiol 2016;45(6):1887–1894. doi: 10.1093/ije/dyw341 [DOI] [PubMed] [Google Scholar]

[R36] 36.Rigby RA, Stasinopoulos DM. Generalized additive models for location, scale and shape. J R Stat Soc Ser C Appl Stat 2005;54(3):507–554. doi: 10.1111/j.1467-9876.2005.00510.x [DOI] [Google Scholar]

[R37] 37.Morgan XC, Tickle TL, Sokol H, et al. Dysfunction of the intestinal microbiome in inflammatory bowel disease and treatment. Genome Biol 2012;13(9):R79. doi: 10.1186/gb-2012-13-9-r79 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R38] 38.Lin H, Peddada SD. Multigroup analysis of compositions of microbiomes with covariate adjustments and repeated measures. Nat Methods. 2024;21(1):83–91. doi: 10.1038/s41592-023-02092-7 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R39] 39.Tingley D, Yamamoto T, Hirose K, Keele L, Imai K. mediation: R Package for Causal Mediation Analysis. J Stat Softw 2014;59:1–38. doi: 10.18637/jss.v059.i0526917999 [DOI] [Google Scholar]

[R40] 40.Glickman ME, Rao SR, Schultz MR. False discovery rate control is a recommended alternative to Bonferroni-type adjustments in health studies. J Clin Epidemiol 2014;67(8):850–857. doi: 10.1016/j.jclinepi.2014.03.012 [DOI] [PubMed] [Google Scholar]

[R41] 41.Roa-Díaz ZM, Wehrli F, Lambrinoudaki I, et al. Early menopause and cardiovascular risk factors: a cross-sectional and longitudinal study. Menopause N Y N 2023;30(6):599–606. doi: 10.1097/GME.0000000000002184 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R42] 42.Mauvais-Jarvis F, Manson JE, Stevenson JC, Fonseca VA. Menopausal Hormone Therapy and Type 2 Diabetes Prevention: Evidence, Mechanisms, and Clinical Implications. Endocr Rev 2017;38(3):173–188. doi: 10.1210/er.2016-1146 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R43] 43.Ruuskanen MO, Erawijantari PP, Havulinna AS, et al. Gut Microbiome Composition Is Predictive of Incident Type 2 Diabetes in a Population Cohort of 5,572 Finnish Adults. Diabetes Care. 2022;45(4):811–818. doi: 10.2337/dc21-2358 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R44] 44.Zhai L, Xiao H, Lin C, et al. Gut microbiota-derived tryptamine and phenethylamine impair insulin sensitivity in metabolic syndrome and irritable bowel syndrome. Nat Commun 2023;14(1):4986. doi: 10.1038/s41467-023-40552-y [DOI] [PMC free article] [PubMed] [Google Scholar]

[R45] 45.Hosomi K, Saito M, Park J, et al. Oral administration of Blautia wexlerae ameliorates obesity and type 2 diabetes via metabolic remodeling of the gut microbiota. Nat Commun 2022;13(1):4477. doi: 10.1038/s41467-022-32015-7 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R46] 46.Takeuchi T, Kubota T, Nakanishi Y, et al. Gut microbial carbohydrate metabolism contributes to insulin resistance. Nature. 2023;621(7978):389–395. doi: 10.1038/s41586-023-06466-x [DOI] [PMC free article] [PubMed] [Google Scholar]

[R47] 47.Depommier C, Everard A, Druart C, et al. Supplementation with Akkermansia muciniphila in overweight and obese human volunteers: a proof-of-concept exploratory study. Nat Med 2019;25(7):1096–1103. doi: 10.1038/s41591-019-0495-2 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R48] 48.Wang Y, Sharma A, Weber KM, et al. The menopause-related gut microbiome: associations with metabolomics, inflammatory protein markers, and cardiometabolic health in women with HIV. Menopause N Y N 2024;31(1):52–64. doi: 10.1097/GME.0000000000002287 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R49] 49.Palmer ND, Okut H, Hsu FC, et al. Metabolomics Identifies Distinctive Metabolite Signatures for Measures of Glucose Homeostasis: The Insulin Resistance Atherosclerosis Family Study (IRAS-FS). J Clin Endocrinol Metab 2018;103(5):1877–1888. doi: 10.1210/jc.2017-02203 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R50] 50.Razquin C, Toledo E, Clish CB, et al. Plasma Lipidomic Profiling and Risk of Type 2 Diabetes in the PREDIMED Trial. Diabetes Care. 2018;41(12):2617–2624. doi: 10.2337/dc18-0840 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R51] 51.Beyene HB, Huynh K, Wang T, et al. Development and validation of a plasmalogen score as an independent modifiable marker of metabolic health: population based observational studies and a placebo-controlled cross-over study. EBioMedicine. 2024;105:105187. doi: 10.1016/j.ebiom.2024.105187 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R52] 52.Jackson DR, Cassilly CD, Plichta DR, et al. Plasmalogen Biosynthesis by Anaerobic Bacteria: Identification of a Two-Gene Operon Responsible for Plasmalogen Production in Clostridium perfringens. ACS Chem Biol 2021;16(1):6–13. doi: 10.1021/acschembio.0c00673 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R53] 53.Mawatari S, Sasuga Y, Morisaki T, Okubo M, Emura T, Fujino T. Identification of plasmalogens in Bifidobacterium longum, but not in Bifidobacterium animalis. Sci Rep 2020;10(1):427. doi: 10.1038/s41598-019-57309-7 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R54] 54.Joyce LR, Guan Z, Palmer KL. Streptococcus pneumoniae, S. pyogenes and S. agalactiae membrane phospholipid remodelling in response to human serum. Microbiol Read Engl 2021;167(5):001048. doi: 10.1099/mic.0.001048 [DOI] [PMC free article] [PubMed] [Google Scholar]

[R55] 55.Mu R, Momeni S, Krieger M, et al. Plasmalogen, a glycerophospholipid crucial for Streptococcus mutans acid tolerance and colonization. Appl Environ Microbiol 2024;90(4):e0150023. doi: 10.1128/aem.01500-23 [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Menopause Factors and Alterations in Gut Microbiota and Insulin Homeostasis: A Cross-Sectional Analysis of the Microbiome and Insulin Longitudinal Evaluation Study (MILES)

Vincent J Maffei, MD PhD

Alain G Bertoni, MD MPH

Alexis C Wood, PhD

Jerome I Rotter, PhD

Osa Crago, MS

Yii-Der I Chen, PhD

Joseph F Petrosino, PhD

Kristi L Hoffman, MPH PhD

Mark O Goodarzi, MD PhD

Elizabeth T Jensen, MPH PhD

Abstract

Aim

Materials and Methods

Results

Conclusions

Introduction

Materials and Methods

Participants

Soluble Fiber Intake and Physical Activity Assessments

Menopause History Assessment

Dysglycemia and Insulin Homeostasis Traits

Fecal Microbiota Taxonomic Profiling

Plasma Metabolite Quantification

Microbiota Feature Selection

Beta Diversity

Microbiota Co-abundance Network

Statistical Analysis

Results

Table 1.

Menopause factors and insulin homeostasis traits

Table 2. Menopause factors are associated with insulin homeostasis traits.

Fecal microbiota composition and pre-menopausal gynecologic surgery

Beta-diversity Analysis

Figure 1. Hysterectomy and/or bilateral oophorectomy before menopause and a menopause age <50 years are associated with fecal microbial beta-diversity.

Co-abundance Network Analysis

Figure 2. Hysterectomy and/or bilateral oophorectomy before menopause is associated with modules of co-abundant fecal microbial species.

Pre-menopausal gynecologic surgery and fecal microbiota-associated plasma metabolites

Figure 3. Hysterectomy and/or bilateral oophorectomy before menopause and a history of hormone replacement therapy are associated with microbiota-associated plasma metabolite levels.

Plasmalogen lipid metabolites and insulin sensitivity

Discussion

Supplementary Material

Acknowledgments

References

Associated Data

Supplementary Materials

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases