Introduction
Women spend more than one-third of their lives in a post-menopausal state. Menopause is the definitive disappearance of menstruation due to the depletion of ovarian activity, occurring after 12 months of amenorrhoea.1Santoro N Roeca C Peters BA Neal-Perry G. The menopause transition: signs, symptoms, and management options. The menopause transition, also known as perimenopause, is the beginning of menstrual irregularities when symptoms of female sex hormone deficiency begin. Menopausal change is associated with higher prevalence of metabolic syndrome and cardiovascular risk factors, as well as alterations in mood, sleep, diet and other lifestyle factors.2Zhu D Chung HF Dobson AJ et al.Type of menopause, age of menopause and variations in the risk of incident cardiovascular disease: pooled analysis of individual data from 10 international studies., 3Christakis MK Hasan H De Souza LR Shirreff L. The effect of menopause on metabolic syndrome: cross-sectional results from the Canadian longitudinal study on., 4Bromberger JT Kravitz HM. Mood and menopause: findings from the Study of Women’s Health Across the Nation (SWAN) over 10 years. However, whether the reported changes associated with menopause are due to hormonal alterations, psychological changes associated with the transition, natural ageing, social and behavioural factors of midlife or genetic vulnerability is less clear and warrants further exploration.Extensive evidence shows that changes in body composition including loss of lean body mass, accumulation of fat mass and redistribution of the adipose tissue in the abdominal area occur with menopause.5Abildgaard J Ploug T Al-Saoudi E et al.Changes in abdominal subcutaneous adipose tissue phenotype following menopause is associated with increased visceral fat mass. Subsequently, unfavourable fasting blood measures3Christakis MK Hasan H De Souza LR Shirreff L. The effect of menopause on metabolic syndrome: cross-sectional results from the Canadian longitudinal study on. and a shift to an atherogenic lipid profile (increases in total cholesterol, low-density lipoprotein cholesterol (LDL-C), and apolipoprotein B) occur independently of age, due to menopause.2Zhu D Chung HF Dobson AJ et al.Type of menopause, age of menopause and variations in the risk of incident cardiovascular disease: pooled analysis of individual data from 10 international studies. However, less is known regarding the impact of menopause on the integrated postprandial metabolic response.6Rathnayake KM Weech M Jackson KG Lovegrove JA. Impact of meal fatty acid composition on postprandial lipaemia, vascular function and blood pressure in postmenopausal women. Given that humans spend the majority of their day in the postprandial (1–8 h post eating) phase (∼18 h/d) and postprandial lipaemia and glycaemia are independent risk factors for cardiovascular diseases (due to their downstream effects on inflammation, oxidative stress, haemostatic function and lipoprotein remodelling), studies exploring multi-factorial postprandial responses with respect to menopausal status are needed. Furthermore, the gut microbiome is increasingly recognised as an important regulator of metabolism and is associated with multiple cardiometabolic risk factors.7Asnicar F Berry SE Valdes AM et al.Microbiome connections with host metabolism and habitual diet from 1,098 deeply phenotyped individuals.,8Menopause, the gut microbiome, and weight gain: correlation or causation?. Whilst alterations in gut microbiome composition have been shown during the menopausal transition, its role in increased metabolic risk faced by menopausal women remains unclear.8Menopause, the gut microbiome, and weight gain: correlation or causation?.
In light of the well-recognised changes that occur in lifestyle and body composition upon menopause, research furthering our understanding of the key metabolic and microbial changes occurring in concert may help provide tailored lifestyle and dietary advice for women during their menopausal transition and post-menopause. This study leveraged the densely phenotyped ZOE PREDICT cohort to, firstly, characterise lifestyle, diet and health measures in pre-, peri- and post-menopausal women and, secondly, explore the physiological changes of menopause with a focus on postprandial metabolism and the gut microbiome. We report; 1) differences in body composition, fasting blood measures, postprandial metabolites, lifestyle, diet, microbiome and mood across sex, age and menopausal status, 2) an independent association of menopause with postprandial glucose responses in an age-matched subgroup, 3) a protective association between menopausal hormone therapy (MHT) use and visceral fat and fasting and postprandial measures, and 4) a mediation effect of diet and bacterial species on visceral fat and inflammation, by menopause status.
Results
The ZOE PREDICT 1 cohort
The ZOE PREDICT 1 study collected detailed phenotypic data, habitual diet information, gut microbiome data, and multiple fasting and postprandial cardiometabolic blood marker measurements from 1002 healthy adults from the United Kingdom (NCT03479866), Figure 1. Postprandial measures were collected in the clinic (0–6 h; serum triglyceride (TG), glucose, insulin, glycoprotein acetylation (GlycA) and circulating metabolite (NMR metabolomics panel) concentrations) following sequential mixed-nutrient dietary challenges and remotely (13d) for glucose, using continuous glucose monitors (CGMs) following 8 standardised meals of differing macronutrient content. Self-reported pre-, peri- and post-menopausal females were selected from PREDICT 1 (n=366, n=55, n=206 respectively). Characteristics related to menopausal status including age at menopause, contraception use, MHT use and smoking status are reported in Supplementary Table 2. The study design and the inclusion criteria of study participants are described in detail in the Methods. Further information on research design has been previously published.9Berry SE Valdes AM Drew DA et al.Human postprandial responses to food and potential for precision nutrition. Compared to the average UK population, PREDICT 1 participants were older (mean age 46 vs 41 years respectively), had a lower BMI (26 vs 28 kg/m2 respectively), were less likely to smoke and the proportion of males was also lower (27% vs 49 % respectively).
Figure 1Experimental design. Self-reported pre-, peri- and post-menopausal females were selected from the ZOE PREDICT 1 (UK; n=366, n=55, n=206 respectively) cohort. Phenotypic data were obtained following in-person assessments and during a 13-d at home phase. Personal characteristics, lifestyle factors, diet, fasting and postprandial metabolic response, test meals, continuous glucose levels, gut microbiome composition and mood, were examined across menopausal groups in the PREDICT 1 cohort.
Cohort characteristics and relationships with sex, age and menopausal status
Differences in the association of age with metabolic traits between males and females
We first characterised the role of age and sex in cardiometabolic health, anthropometry, lifestyle and diet measures (Table 1). Males had more unfavourable body composition, fasting and postprandial blood profiles and a lower diet quality than females (pr=0.35, GlycA (a measure of inflammation); r=0.09, TG; r=0.11, insulin; r=−0·09, LDL-C; r=0.42, glucose2hiauc; r=0.21 and triglyceride6hiauc; r=0.22), whereas the opposite was seen for liver probability scores, which estimates the presence of fatty liver (males; r=−0.16 vs. females; r=0.08) (Figure 2 a-e). These relationships persisted in a subcohort of females (apart from TG, insulin and GlycA), matched in sample size (n=274, mean of multiple random sub samples reported) to the male cohort.
Figure 2Variation in males and females across age groups. Age groups selected to represent the pre-menopausal period (18–32y and 33–43y, pre-menopausal females only), the menopausal transition period (44–58y, includes pre-, peri- and post-menopausal females) and the post-menopausal period (>58y, post-menopausal females only) a. Systolic BP (mm/Hg) b. LDL-C (mmol/L), c. Glucose2hiauc, d. Triglyceride6hiauc, e. Liver probability score. Black dots and lines represent males (N=273; 18–32y, n=68; 33–43y, n=64; 44-58y, n=110; >58y, n=32), green dots and lines represent females (N=725; 18–32y, n=111; 33–43y, n=153; 44-58y, n=339; >58y, n=122). Variation in clinic postprandial metabolic responses for f. Glucose, g. TG, h. Insulin and i. GlycA in pre- (n=365), peri- (n=55) and post-menopausal (n=206) females. Lines represent average values for pre- (black), peri- (orange) and post-menopausal (blue) females. j. CGM-derived peak glucose (0–2 h) concentrations following 8 set meals (metabolic challenge, high fibre, high fat, high protein, high carbohydrate, oral glucose tolerance test, US/UK representative (breakfast and lunch) meals) for pre (n=365) and post-menopausal (n=206) females. k. Gut bacterial species with significantly different relative abundances (n species=8) between pre- vs. post-menopause females (n=564) in the PREDICT 1 cohort.
Associations with menopausal status
Given the divergent age-sex response, we then characterised the female cohort according to pre-, peri- and post-menopausal status (adjusting for age, BMI, MHT use and smoking status). Following the menopause transition, post-menopausal females were significantly older (mean difference 19.9 years, 95% CI; 18.7–21.0), had higher SBP (12.2 mm/Hg, 95% CI; 9.6–14.8), had unfavourable fasting blood concentrations (higher glucose; 0.30 mmol/L, 95% CI; 0.23–0.39, insulin; 0.44 mmol/L, 95% CI; 0.30–1.19, GlycA; 0.07 mmol/L, 95% CI; 0.05–0.10 and glycosylated haemoglobin (HbA1c); 0.22%, 95% CI; 0.18–0.20), higher insulin insensitivity (HOMA-IR; 0.20, 95% CI; 0.01–0.40) and higher ASCVD 10y risk (0.02, 95% CI; 0.02–0.03) (Table 2). No significant differences were observed in fasting lipoprotein size, concentration or composition from NMR metabolomic analysis after FDR adjustment (Supplementary Table 3). Post-menopausal females also reported greater sleep difficulties (higher Pittsburgh Sleep Quality Index; ppvs. pre-menopausal females it was not statistically significant. Visceral fat mass, whole body fat percentage and bone mineral density, as measured by DEXA, were also not different between groups. Differences remained in the same direction between pre- and post-menopausal females when stratified according to BMI; Supplementary Table 5.
Table 2Characteristics of the PREDICT 1 cohort across menopausal status.
*Scoring ranges from lowest “all of the time” to highest “none of the time”. ANCOVA (continuous) and logistic regression (categorical) adjusted for age and BMI. Anthropometric traits were adjusted for age only. Differences between menopausal groups also adjusted for MHT use and smoking status. All p-values adjusted for multiple testing (FDR<0.05); * p<0.05, ** p<0.01, ***p<0.001.+ PREDICT 1 females who self-reported menopausal status.
Menopausal status associations with postprandial responses and glycaemic variability
Postprandial elevations in circulating lipids and glucose are associated with increased risk of cardiometabolic disease, type-2 diabetes and obesity, independent from fasting measures.9Berry SE Valdes AM Drew DA et al.Human postprandial responses to food and potential for precision nutrition. Therefore, as humans spend the majority of their day in the postprandial metabolic state, we next examined differences in postprandial responses between pre- and post-menopausal females (adjusting for age, BMI, MHT use and smoking status). Clinic postprandial metabolic responses differed between groups (Figure 2 f-i), with significantly higher glucose2hiauc and insulin2hiauc (pvs. pre-menopausal females.
Given the differences in postprandial glycaemia measured in the tightly controlled clinic setting, we then explored glycaemia in the remote phase of the study using CGM data. We examined different features of glycaemic responses, including glycaemic variability (measured by coefficient of variation), time spent in range, mean day long glucose concentration, as well as glucose2hiauc and peak0-2h following meals of varying macronutrient composition. Mean day-long glucose concentrations and glycaemic variability (examined using 2-4 free-living days of the PREDICT 1 remote phase) were higher in post-menopausal females (5.1±0.53 mmol/L and 17.6±4.3 %) compared to pre-menopausal females (4.9±0.54 mmol/L and 15.6±4.00 %), p<0.002 (ANCOVA) (Supplementary Table 6). Pre-menopausal females also elicited a more favourable TIR (3.9–5.6 mmol/L) (70.8%±16.9) compared with post-menopausal females (68.8%±15.6), p<0.05 (ANCOVA).
Glucose2hiauc and glucose peak0-2h were next examined across seven isocaloric meals differing in protein, fat, carbohydrate, and fibre content, as well as an oral glucose tolerance test (OGTT) following an overnight fast (nutritional composition of meals available in Supplementary Table 1). For all meals, postprandial glucose concentrations (glucose2hiauc and peak0-2h) were higher in post- vs. pre-menopausal females (Supplementary Table 7), with the greatest differences observed following a nutrient composition reflective of a typical UK/US average diet (composition 40% fat and 57% carbohydrate; mean difference in glucose2hiauc 3524.3 and peak0-2h 0.89 mmol/L, p2hiauc 2598.2 and peak0-2h 0.80 mmol/L) and high carbohydrate meal (glucose2hiauc 2291.5 and peak0-2h 0.79 mmol/L). Glycaemic responses (glucose2hiauc) remained significantly higher in post-menopausal females for the typical UK/US average meal and (peak0-2h) for high carbohydrate meal, typical UK/US average meals and OGTT, after correcting for covariates compared with pre-menopausal females (pFigure 2j and Supplementary Table 7). Postprandial lipoprotein profiles (NMR analysis) were not significantly different (FDR adjusted).
Menopausal status also corresponded to a state of greater inter-individual variability (coefficient of variation) in post- vs. pre-menopausal females for postprandial insulin (30 min, pre-menopause 89% vs. post-menopause 200%) and HOMA-IR (pre-menopause 69.4% vs. post-menopause 82.6%) (Supplementary Figure 1 and Supplementary Table 8).
Menopausal status associated with microbiome composition
We have previously reported a strong association between gut microbiome composition and metabolic health (both fasting and postprandial measures).7Asnicar F Berry SE Valdes AM et al.Microbiome connections with host metabolism and habitual diet from 1,098 deeply phenotyped individuals. Therefore, we examined gut microbiome composition, within-sample richness (number of species) and within-sample diversity (Shannon) in pre- and post-menopausal females (n=564). The relative abundances of microbiome species differed with menopausal status, with eight species significantly differentially abundant after correction for multiple testing (pFigure 2k). Of these, four species had significantly higher abundances in pre- vs. post-menopausal females, whereas four species had higher abundances post-menopause (pBacteroides ovatus has been associated with younger age in a large meta-analysis,12
Manghi P., Schiffer L., Golzato D., et al (in press) Meta-analysis of 20,533 human metagenomes defines an index of oral to gut microbial introgression and associations with age, sex, BMI, and diseases. Nature Metabolism.
but none of the other species were associated with age or BMI. Microbiome richness and diversity were not significantly different (Supplementary Table 9).
Is the association of menopause with metabolic traits independent of age?
Menopausal status is an age-related biological event; thus to untangle some of the effects of age from menopausal status we examined cardiometabolic health, anthropometry, lifestyle and diet measures in an age-matched subgroup of pre- and post-menopausal females. Measures that were significantly different with menopausal status in the total cohort were examined (adjusted for age, BMI, MHT use and smoking status). Sleep and intakes of sugar remained significantly different in the age-matched subgroup (n=150, age range 47–56y) (Figure 3a-b) (Supplementary Table 10). Postprandial glycaemic measures also remained significantly higher in post- vs. pre-menopausal females (p0-2h (7.62±1.15 mmol/L vs. 7.19±1.03 mmol/L (Figure 3c)); CGM glycaemic variability (17.9±4.1 % vs. 15.9±4.4 %) (Figure 3d) and glucose2hiauc (150558±5384 vs. 13258±5657; following meal reflective of a typical UK/US average diet) (Figure 3e).
Figure 3Distribution of a. sugar intake, b. sleep quality, c. postprandial glucose peak0-2h and d. glycaemic variability between age-matched pre- (n=86) and post-menopausal (n=64) females. The red line represents the group median. e. Variation in postprandial glucose responses (0–2 h) to a typical UK/US average lunch, measured by CGM. Pre-m; pre-menopausal, Post-m; post-menopausal.
When investigating differences in microbiome composition, in the age matched cohort, the abundances of six species were significantly different between pre- and post-menopausal females but were not significant after FDR correction (Supplementary Table 9). From the eight species previously identified in the total cohort, four species showed the same directional trend but were not significant after FDR correction (Eubacterium hallii, Bifidobacterium adolescentis, Faecalibacterium prausnitzii, Oscillibacter sp PC13). The remaining four species, (Harryflintia acetispora, Bacteroides ovatus, Lawsonibacter asaccharolyticus and Clostridium disporicum) were not abundant in the same directional trend in the age-matched subgroup.
Due to the deficiency in female sexual hormones observed post-menopause we also examined measures in age-matched 1) males, 2) pre-menopausal females and 3) post-menopausal females (age range 47–56y, Supplementary Table 10). Females, both pre- and post-menopausal, had significantly lower SBP and ASCVD 10y risk compared to males. However, pre-menopause was associated with more favourable fasting blood glucose and GlycA concentrations, and mean day long glucose concentrations (CGM) compared to males (p<0.05) (ANCOVA), while male levels were more similar to post-menopausal females. Interestingly, glycaemic variability, which worsened post-menopause (compared with age-matched pre-menopausal females), was also significantly higher than males (p=0.007) (ANCOVA), suggesting unfavourable blood glucose variability independent of age and gender.
The association between post-menopausal hormone therapy use and metabolic health
MHT is commonly used among females undergoing the menopausal transition. Given the replacement of vital female sexual hormones, most notably oestrogen, by MHT and the involvement of these in gut-metabolic interactions,13Vieira AT Castelo PM Ribeiro DA Ferreira CM Influence of oral and gut microbiota in the health of menopausal women. we examined the link between MHT use and metabolic health in post-menopausal females (Table 3; adjusted for age, BMI and smoking status). Measures that were significantly different with menopausal status in the total cohort were examined along with a selection of measures with previously known associations with MHT (visceral fat and lipids (LDL-C and triglyceride6hiauc)).14Salpeter SR Walsh JM Ormiston TM Greyber E Buckley NS Salpeter EE. Meta-analysis: effect of hormone-replacement therapy on components of the metabolic syndrome in postmenopausal women. Post-menopausal females using MHT (n=35) had lower visceral fat mass, favourable fasting blood concentrations (for glucose, insulin, cholesterol (total and LDL)), and lower postprandial lipaemia (triglyceride6hiauc) compared to non-MHT users (n=172), pTable 3). Furthermore, to disentangle the effect of genetics, we examined our predominantly twin cohort and identified six post-menopausal MZ twin pairs discordant for MHT use. HbA1c showed significant differences between discordant twin pairs (p=0.004) (Paired t-test) (Supplementary Table 12). We observed a more favourable body composition (lower BMI, weight, visceral fat) and blood biomarker concentrations (lower glucose, insulin, TG and GlycA) in twins using MHT vs. those non-MHT users (although not significant).
Table 3Menopausal hormone therapy use in post-menopausal females.
*Post-menopausal females who self-reported currently taking MHT.
Mediating effects of sleep, physical activity, diet and microbiome on the link between menopausal status and key metabolic health indicators
Changes to lifestyle, diet, mood, anthropometry, gut and cardiometabolic health following the menopause transition are intricately linked. We hypothesised that key pillars of health, including sleep, physical activity, diet and gut health, may mediate the adverse metabolic effects observed post-menopause for several health outcomes. Thus we conducted a formal mediation analysis to determine the mediating effect of sleep, physical activity, diet quality (Healthy Eating Index (HEI)) and microbiome species to assess the link between menopausal status and key metabolic health indicators (visceral fat, inflammation (GlycA360mins) and glycaemia (peak0-2h)). Diet quality, in part, mediated the association between menopause status and visceral fat (proportion mediated; 9%, pFigure 4a and Supplementary Table 13). For the microbiome species, we examined the eight species that were different between pre- and post-menopausal females in the total cohort (Figure 2k) and the 30 species previously associated with cardiometabolic health and diet in this cohort.7Asnicar F Berry SE Valdes AM et al.Microbiome connections with host metabolism and habitual diet from 1,098 deeply phenotyped individuals. Four microbiome species (Collinsella intestinalis, Eggerthella lenta, Flavonifractor plautii and Ruminococcus gnavus) acted as partial mediators in the association between menopause status and GlycA (proportion mediated; 5–10%) (Figure 4b and Supplementary Table 13). The species Flavonifractor plautii and Eggerthella lenta acted as a partial mediators in the association between menopause status and visceral fat (proportion mediated; 5%) and glucose (peak0-2h) (proportion mediated; 8%, pOscillibacter sp 57 20 acted as partial mediators in the association between menopause status and glucose (peak0-2h) (proportion mediated; −6%) (Figure 4c) (p
Figure 4Mediation analysis of the association between menopausal status and a. visceral fat (g) b. glucose (peak0-2h) c. GlycA (mmol/L). Average direct effect (ADE) and average causal mediation effects (ACME) are reported (*P<0.05, **P < 0.01, ***P < 0.001). d. Health and diet measures associated with post-menopausal status.
Discussion
With our ageing population, it is estimated that worldwide 1.2 billion women will be in the menopausal transition or post-menopause by the year 2030.15WHO
Research on the menopause in the 1990s. In the current study, we demonstrate that post-menopause status is associated with unfavourable changes in body composition, fasting and postprandial blood profiles (including inflammation and postprandial glycaemia), diet, sleep and gut microbiome. We further explored the independent role of menopause from age and observed poorer sleep and diet, as well as higher postprandial glycaemic measures post-menopause, alongside a protective association of MHT use with visceral fat, fasting (glucose and insulin) and postprandial (triglyceride6hiauc) blood measures. Finally, we investigated the association between modifiable risk factors on metabolic changes in menopause, finding a mediating effect of diet and a gut bacterial species and visceral fat, glycaemia and inflammation, by menopause status. Changes in key metabolic health indicators observed in menopause may therefore be attenuated by targeting the gut microbiome and diet.Differences in typically measured features of the postprandial responses for glucose and TG have previously been reported in pre- versus post-menopausal females in small studies looking at single postprandial measures.6Rathnayake KM Weech M Jackson KG Lovegrove JA. Impact of meal fatty acid composition on postprandial lipaemia, vascular function and blood pressure in postmenopausal women.,16The impact of years since menopause on the development of impaired glucose tolerance.,17Jackson KG Abraham EC Smith AM et al.Impact of age and menopausal status on the postprandial triacylglycerol response in healthy women. Here, we show that in addition to 2h iAUC for glucose and insulin, post-menopausal females had a more unfavourable glycaemic variability, TIR and day-long glucose, measured by CGM in free-living days. Continuous glycaemic features capture day-to-day glycaemic excursions, including peak concentration, nadirs ‘below baseline’, glycaemic variability and TIR, each associated with downstream metabolic effects, including oxidative stress, inflammation, and increased cardiovascular and diabetes disease risk.18Glycemic variability: how do we measure it and why is it important?. In this cohort, we did not see differences in postprandial TG independent of age. To the best of our knowledge, one previous study has compared postprandial TG response between pre- and post-menopausal females using a sequential meal postprandial investigation.17Jackson KG Abraham EC Smith AM et al.Impact of age and menopausal status on the postprandial triacylglycerol response in healthy women. Females were subdivided into both younger and older pre- and post-menopause subgroups but no differences in postprandial TG were observed due to menopausal status.17Jackson KG Abraham EC Smith AM et al.Impact of age and menopausal status on the postprandial triacylglycerol response in healthy women. Differences were observed between young and old pre-menopausal groups, suggestive that major increases observed in postprandial TG may occur during the later pre-menopause years.17Jackson KG Abraham EC Smith AM et al.Impact of age and menopausal status on the postprandial triacylglycerol response in healthy women. This study highlights that inherent biological differences exist between males and females while also demonstrating the protective effects of oestrogen through the controlled comparison of age-matched males and pre- and post-menopausal females. Multiple aspects of health and glucose homeostasis are regulated differently between sexes, and our findings highlight the added complexity introduced by the female menopausal transition.The protective association of MHT use post-menopause with metabolic measures, further supports an independent effect of oestrogen observed in existing research. Favourable effects on visceral fat mass and fasting blood concentrations and lower postprandial lipaemia post-menopause are in accordance with evidence that MHT can provide effective relief in low-risk females for a wide range of adverse health outcomes and symptoms associated with menopause.14Salpeter SR Walsh JM Ormiston TM Greyber E Buckley NS Salpeter EE. Meta-analysis: effect of hormone-replacement therapy on components of the metabolic syndrome in postmenopausal women.,19Langer RD Hodis HN Lobo RA Allison MA. Hormone replacement therapy – where are we now?. However, MHT is available in multiple forms and doses, and the impact of different therapies is beyond the scope of this research. For example, transdermal oestradiol and micronised progesterone are not associated with a risk of venous thromboembolism compared to oral oestrogen with a synthetic progestogen.20Postmenopausal hormone replacement therapy and cardiovascular disease: the value of transdermal estradiol and micronized progesterone. Given the appropriate dose and type, MHT can reduce biological ageing and provide effective treatment to alleviate menopausal symptoms and confer protective cardiometabolic effects in appropriate candidates.20Postmenopausal hormone replacement therapy and cardiovascular disease: the value of transdermal estradiol and micronized progesterone.Female sex hormones influence the microbiota at multiple body sites, including the gut,13Vieira AT Castelo PM Ribeiro DA Ferreira CM Influence of oral and gut microbiota in the health of menopausal women. which has been associated with multiple diseases outside of this organ. The gut microbiota metabolizes oestrogen-like compounds consumed in plant foods (phytoestrogens), including lignans (derived from a variety of plant foods) and isoflavonoids (found in soy foods).21Frankenfeld CL Atkinson C Wähälä K Lampe JW. Obesity prevalence in relation to gut microbial environments capable of producing equol or O-desmethylangolensin from the isoflavone daidzein.,22Nakatsu CH Armstrong A Clavijo AP Martin BR Barnes S Weaver CM. Fecal bacterial community changes associated with isoflavone metabolites in postmenopausal women after soy bar consumption. Administration of isoflavone-rich soy foods to post-menopausal females can lead to elevated levels of gut microbial derived oestrogen-like metabolites21Frankenfeld CL Atkinson C Wähälä K Lampe JW. Obesity prevalence in relation to gut microbial environments capable of producing equol or O-desmethylangolensin from the isoflavone daidzein. and changes in some bacteria including increases in Bifidobacterium and decreases of Clostridiaceae, which play roles in inflammatory diseases.22Nakatsu CH Armstrong A Clavijo AP Martin BR Barnes S Weaver CM. Fecal bacterial community changes associated with isoflavone metabolites in postmenopausal women after soy bar consumption. Further, changes in oestrogen receptor (ER-β) expression have been shown to affect microbiota composition23Menon R Watson SE Thomas LN et al.Diet complexity and estrogen receptor β status affect the composition of the murine intestinal microbiota. and the gut microbiomes of pregnant females were profoundly altered during the third trimester, when oestrogen is at its peak.24Koren O Goodrich JK Cullender TC et al.Host remodeling of the gut microbiome and metabolic changes during pregnancy. Thus a reciprocal relationship may exist between oestrogen and the microbiome which may modulate the health of menopausal females. Our findings show differences in abundances of species post-menopause, including pro-inflammatory and obesogenic bacteria. Of interest, four species in part modulated the relationship between menopause and GlycA, a marker of inflammation. Our previous research associated these species with unfavourable cardiometabolic health, diet and inflammatory outcomes,7Asnicar F Berry SE Valdes AM et al.Microbiome connections with host metabolism and habitual diet from 1,098 deeply phenotyped individuals. in line with previous research.25Henke MT Kenny DJ Cassilly CD Vlamakis H Xavier RJ Clardy J. Ruminococcus gnavus, a member of the human gut microbiome associated with Crohn’s disease, produces an inflammatory polysaccharide., 26Thota VR Dacha S Natarajan A Nerad J. Eggerthella lenta bacteremia in a Crohn’s disease patient after ileocecal resection., 27Gupta A Dhakan DB Maji A et al.Association of flavonifractor plautii, a flavonoid-degrading bacterium, with the gut microbiome of colorectal cancer patients in India. For example, Ruminococcus gnavus, a prevalent gut microbe, produces a potent, inflammatory polysaccharide recognized by innate immune cells through the toll-like receptor 4 (TLR4) and is associated with multiple inflammatory diseases.25Henke MT Kenny DJ Cassilly CD Vlamakis H Xavier RJ Clardy J. Ruminococcus gnavus, a member of the human gut microbiome associated with Crohn’s disease, produces an inflammatory polysaccharide. GlycA is strongly associated with cardiovascular and diabetes risk28Mazidi M Valdes AM Ordovas JM et al.Meal-induced inflammation: postprandial insights from the Personalised REsponses to DIetary Composition Trial (PREDICT) study in 1000 participants.; thus, our data suggests that inflammation may be reduced through intervention aimed at improving the gut microbiome post-menopause.The mediating effects of diet and microbiome on visceral fat, glyceamia and inflammation respectively, suggest that modifiable factors may play a role in the unfavourable changes observed post-menopause. Given that many diet and lifestyle changes also occur during the menopausal transition,29ESHRE Capri Workshop Group
Perimenopausal risk factors and future health. which are known to impact metabolic health outcomes, these are potential targets to alleviate some of the downstream unfavourable health effects associated with menopause. For example, in our cohort, post-menopausal females consumed higher intakes of dietary sugars and reported poorer sleep. These are both associated with more pronounced postprandial glycaemia30Tsereteli N Vallat R Fernandez-Tajes J et al.Impact of insufficient sleep on dysregulated blood glucose control under standardised meal conditions. and increased risk of type-2 diabetes and cardiovascular disease.31O’Connor L Imamura F Brage S Griffin SJ Wareham NJ Forouhi NG. Intakes and sources of dietary sugars and their association with metabolic and inflammatory markers. Further, a decrease in physical activity energy expenditure and a shift to a more sedentary lifestyle associated with menopause32Duval K Prud’homme D Rabasa-Lhoret R et al.Effects of the menopausal transition on energy expenditure: a MONET Group Study. was recently demonstrated to be a direct effect of declining oestrogen33Krause WC Rodriguez R Gegenhuber B et al.Oestrogen engages brain MC4R signalling to drive physical activity in female mice. which may have been captured with more quantitative measures of physical activity in this study. These observed changes in diet and physical activity may increase the risk for positive energy balance and weight gain over time. Healthy dietary patterns such as the Mediterranean diet have been associated with improved weight and vasomotor symptoms in post-menopausal females34Herber-Gast GC Mishra GD. Fruit, Mediterranean-style, and high-fat and -sugar diets are associated with the risk of night sweats and hot flushes in midlife: results from a prospective cohort study. and certain foods have also been linked to later onset of menopause i.e. intakes of green and yellow vegetables as well as fresh legumes.35Dunneram Y Greenwood DC Cade JE. Diet, menopause and the risk of ovarian, endometrial and breast cancer. Positive health effects associated with diet quality, may be due to higher contents of dietary fiber, complex carbohydrates, vitamins, minerals, polyunsaturated fatty acids, and phytochemicals. Diets high in plant-based foods may be naturally rich in isoflavones which may play a role in the protective effects some diet patterns such as Mediterranean diet have on menopause. Research shows that 60% of women consulted healthcare providers during their menopausal transition seeking support for menopausal symptoms and treatment,36Williams RE Kalilani L DiBenedetti DB Zhou X Fehnel SE Clark RV. Healthcare seeking and treatment for menopausal symptoms in the United States. highlighting an opportune window to target modifiable factors including diet and lifestyle as well as considering MHT.
Limitations of this cross-sectional analysis include; 1) potential inaccuracy in the self-reported time since menopause and self-identification of menopausal status due to ambiguity in determining amenorrhea; 2) cross-sectional data preventing identification of causal relationships; 3) inability to fully account for age-related effects despite the creation of an age-matched subgroup and assessment of MHT discordance; 4) inability to determine menstrual cycle regularity, current contraception use, MHT type (transdermal vs. oral) and other current medication use. However, the data presented links changes in postprandial metabolism, metabolic syndrome factors, mood, sleep, diet and the gut microbiome in a single deeply phenotyped cohort. This may help inform specific hypotheses to design dietary intervention studies examining the impact of menopause status on postprandial metabolism and microbiome composition.
In summary, the physiological effects of menopause are numerous and the menopause transition is a time of great change and unfavourable metabolic effects. Whilst this transition is inevitable, this analysis demonstrates that approaches can be taken to attenuate the adverse cardiometabolic sequelae, including a focus on modifiable factors, such as diet, microbiome and use of MHT in appropriate candidates.
Contributors
Study design and developed concept: S.E.B, A.M.V, J.W, G.H, R.D, N.S, P.W.F, T.D.S. Data collection: S.E.B, I.L, J.W, G.H, T.D.S. Data analysis: K.M.B, F.A. Study coordination: S.E.B, J.W, I.L, G.H, T.D.S. Writing the manuscript: K.M.B, S.E.B, A.M.V, W.L.H, I.L, K.K, J.W, J.E.M, L.R.N, L.D, J.M.O, A.T.C, T.D.S. Obtained funding: J.W, G.H, T.D.S. Accessed and verified the data: K.M.B, F.A, S.E.B, A.M.V. Decided to submit the manuscript: S.E.B and K.M.B. All authors read and approved the final manuscript.
Declaration of interests
TDS, JW and GH are co-founders of ZOE Ltd (ZOE). AMV, FA, LMD, NS, PWF, SEB, TDS receive payments as consultants to ZOE. GH, IL, JW, KK, RD are employed by ZOE. AMV, GH, IL, JW, KK, LMD, NS, PWF, RD, SEB, TDS also receive options in ZOE. Other authors have no conflict of interest to declare.
Source: news.google.com