Late-Onset Menopause Attenuates Aortic Stiffness in the Postmenopausal Period.

All clinical procedures were approved by the University of Colorado Boulder Institutional Review Board and completed at the Clinical Translational Research Center (CTRC) at the University of Colorado Boulder. Written informed consent was obtained, and the intent, purpose, risks and benefits of the measures were explained to all participants, in accordance with the Declaration of Helsinki. All preclinical procedures were approved by the University of Colorado Boulder Institutional Animal Care and Use Committee and adhered to the guidelines set forth by the National Institutes of Health’s Guide for the Care and Use of Laboratory Animals 35 . Raw data and detailed methods for this study are available upon reasonable request from the corresponding author. Participants were previously enrolled in studies (2014-2025) in the Integrative Physiology of Aging Laboratory at the University of Colorado Boulder. Exclusion criteria for these studies included alcohol dependence, uncontrolled thyroid disease, resting BP >160 mmHg systolic or >100 mmHg diastolic, stage III obesity (body mass index [BMI] >40 kg/m 2 ), or a >2.5 kg change in body mass or medication change within the preceding 3 months. Exclusion criteria for the current analysis were participant report of CVD, diabetes, chronic kidney disease, cancer, chronic obstructive pulmonary disease or cognitive impairment/dementia assessed by questionnaire and verified during a review of medical history, blood chemistry, and medication use and a physical exam by a physician during in person screening; a mini mental state exam (score >24 for absence of mild cognitive impairment) 36 and a resting 12-lead ECG were also conducted. Absence of evidence of atherosclerotic plaques assessed via carotid artery ultrasonography was an inclusion criterion. In addition, a subset of participants (n=71 total; n=24 of 40 (60%) total late-onset PMW, n=47 of 86 (55%) total normal-onset PMW) underwent a physician supervised graded exercise test with 12-lead ECG to confirm absence of coronary artery disease. Current use of hormone therapy (estrogen, testosterone, progesterone and/or a combination) or any use of hormone therapy in the 6 months prior to data collection, an unknown age of final menstrual cycle or menopause status, premature/earlier menopause (before age 45) or a self-reported pre- or peri-menopausal status were also exclusion criteria for this analysis. Women who self-reported a cessation of menses <5 years prior to data collection or who were <55 years of age at the time of data collection were removed from the analysis to minimize any potential acute effects of the menopause transition on study outcomes 37 . Late-onset PMW were identified as those who completed menopause ≥55 years of age and normal-onset PMW were identified as those who completed menopause at age 45-54 years 5 . A subset of PMW in the current analysis were included in an ancillary analysis on endothelial function 26 . Resting systolic and diastolic BP were measured in triplicate over the brachial artery in a seated position after 5 minutes of rest using a semi-automated device (Dinamap XL, Johnson & Johnson). Hypertension was defined as having systolic BP ≥130 mmHg or diastolic BP ≥80 mmHg or use of an anti-hypertensive drug 38 . BMI was assessed using anthropometry 39 and obesity was defined by the NIH clinical guidelines 40 . Leisure physical activity level (MET hours/week) was measured using the Modifiable Activity Questionnaire (MAQ) 41 . Race, ethnicity and current medication use were self-reported during screening. Fasting serum creatinine was measured to calculate eGFR using the Chronic Kidney Disease Epidemiology Collaboration equation (2021 CKD-EPI, race free) 42 . Fasting serum lipids (total cholesterol, high density lipoprotein [HDL] cholesterol, low-density lipoprotein [LDL] cholesterol, and triglycerides) and plasma glucose were also assessed. Dyslipidemia was defined as serum total cholesterol >200 mg/dL, LDL cholesterol >100 mg/dL, HDL cholesterol 150 mg/dL or use of a cholesterol-lowering medication 43 . A questionnaire was administered to assess gynecological history including select factors related to menstrual cycle regularity, menopause, parity and history of female-specific conditions 26 , 44 . Years since menopause was calculated based on current age and self-reported age at menopause using the Stages of Reproductive Aging Workshop +10 Guidelines 45 . To confirm menopause status and estrogen deficiency, the University of Colorado Anschutz Colorado Clinical and Translational Sciences Institute CTRC Core Laboratory analyzed serum follicle-stimulating hormone (FSH) and estradiol (chemiluminescence; Beckman Coulter) in a subset of participants with available serum 26 . Total aortic stiffness was assessed as PWV CF as previously described 11 , 23 . Briefly, applanation tonometry was used to obtain consecutive arterial pressure waveforms from the carotid and femoral arteries (Non-Invasive Hemodynamics Workstation; Cardiovascular Engineering Inc.). Electrocardiogram gating of the R wave was used to determine the time delay between the foot of the pressure waves between the carotid and femoral arteries. PWV CF was calculated as the distance between the artery measurement sites divided by the time delay of the pressure waves between sites 11 , 23 . Participant-specific exponential models were used to calculate structural- (PWV CF,structural ) and load-dependent pulse wave velocity (PWV CF,load ) based on the total aortic stiffness assessed (PWV CF ) 12 , 31 - 34 . Arterial mechanics were characterized by incorporating a nonlinear stiffness parameter at the common reference BP of 120/80 mmHg, which was used to compare PWV CF,structural at a uniform BP across all PMW 12 , 31 - 34 . PWV CF,load was derived by subtracting PWV CF,structural from PWV CF . The contribution of PWV CF,structural and PWV CF,load to total PWV CF was determined by calculating the difference of each component between groups from the difference in PWV CF as a percentage (refer to Supplement ). Equations can be found in the Supplemental Methods . To gain initial insight into whether the circulating milieu of late-onset PMW contributes to differences in structural aortic stiffness between late-onset and normal-onset PMW, we utilized a mouse artery serum exposure bioassay recently developed by our laboratory 20 . Young (3-6 month) female C57BL/6J mice were euthanized and thoracic aortas were excised, rinsed with cold physiological salt solution (cPSS) and cleared of perivascular adipose and connective tissue. Aorta segments were cut (~1 mm in length) and incubated (in duplicate) with 5% (10 μL) serum from late-onset and normal-onset PMW and Dulbecco’s Modified Eagle Medium (DMEM) (190 μL) for 48 hours (5% CO 2 , 37°C) 20 , 47 . To determine the contribution of mitoROS-related oxidative stress to elastic modulus, another set of aortic rings were co-incubated with serum from late-onset or normal-onset PMW and the mitochondria-targeted antioxidant MitoQ (3.5 μM; MitoQ Ltd., Auckland, New Zealand). After incubation, aortas were frozen in cPSS and stored at −80°C. This approach does not significantly impact structural components and mechanical properties of the tissue 20 , 47 - 52 . Later, aortas were thawed and aortic elastic modulus was assessed via pin myography 13 , 20 , 47 , 52 - 55 . The elastic modulus was determined as the slope of the linear regression fit to the final four points of the stress-strain curve 13 , 20 , 47 , 52 - 55 ( Supplemental Methods ). Lipidomics analyses of serum samples from late- and normal-onset PMW were conducted at the University of Colorado School of Medicine Metabolomics Core using ultra-high pressure liquid chromatography-mass spectrometry as previously described 26 (see Supplemental Methods ). To determine whether select lipid metabolites that differed between late- and normal-onset PMW contribute to differences in elastic modulus, aortic rings were co-incubated with DMEM and the lipid metabolite with the greatest fold change between groups at levels corresponding to the absolute difference in serum previously assessed in a cohort of late- and normal-onset PMW 26 . Due to constraints of commercial availability, TG(16:0_18:2_18:0) (abbreviated as TG[16:0]) was used in these experiments as the most structurally similar analog to the specific TG identified to be different between groups. To assess whether TG(16:0) is responsible for inducing mitoROS-related increases in aortic elastic modulus, TG(16:0) was co-incubated with MitoQ and exposed to aortic rings for 48 hours; elastic modulus was assessed 13 , 20 , 47 , 52 - 55 . An unpaired student’s t-test (2-tailed) was used to assess control measures, aortic stiffness, and ex vivo aortic elastic modulus between groups (continuous variables). A Fisher’s Exact or Chi-squared test compared categorical variables between groups. Multiple linear regression models determined the association of age at menopause (continuous variable and categorical variable) with total aortic stiffness adjusted for chronological age and education status (Model 1); model 1 + body mass index, LDL cholesterol, triglycerides and plasma glucose (Model 2); model 2 + physical activity, prior hormone therapy use, and current antihypertensive, cholesterol lowering, antidepressive/anxiety and thyroid medication use (Model 3); data are presented as a β-estimate (95% confidence interval [CI]). An ANCOVA assessed the influence of years since menopause on aortic stiffness and the effects of cardiometabolic factors on elastic modulus in the subset of participants used for the ex vivo assays. Subgroup analyses were conducted to determine whether differences in total aortic stiffness were still present when: only assessing PMW of normal BMI (18-24.9kg/m2), using a normal-onset reference group defined as menopause at age 50-52 years of age, and assessing the data based on prior hormone therapy use 46 . A 2-way ANOVA evaluated the influence of prior hormone therapy use on aortic stiffness and the effects of MitoQ co-incubation with PMW serum and TG(16:0) incubation on ex vivo aortic elastic modulus with Fisher’s LSD post-hoc test. Data are expressed as mean±SEM unless otherwise noted. Statistical significance was set a priori at α≤0.05. Additional statistical information can be found in the Supplemental Methods .

A total of 126 PMW were identified for this analysis, with 40 late-onset PMW and 86 normal-onset PMW. Late- and normal-onset PMW were generally well-matched in terms of chronological age and education, physical activity levels, cholesterol profiles, blood glucose, estradiol and FSH and prevalence of dyslipidemia and thyroid disease ( Table 1 , S1 ). However, compared with normal-onset PMW, late-onset PMW had lower BP and prevalence of hypertension, although a similar proportion of late- and normal-onset PMW were on anti-hypertensive medications ( Tables 1 , S1 ). Moreover, other risk factors such as triglycerides and BMI were higher in normal-onset PMW ( Table 1 ), in line with the greater proportion of PMW in this group classified as obese and one participant diagnosed with metabolic dysfunction-associated steatotic liver disease ( Table S1 ). Because of the similar chronological age between groups, late-onset PMW had a shorter time since menopause (13±5 years) compared with normal-onset PMW (17±6 years) ( Table 1 ). Late-onset PMW also had a longer reproductive lifespan, which was likely driven by their later menopausal age given that age at menarche was not different between groups ( Table 1 ). No other notable differences were observed in female-specific factors. Our first goal was to assess if aortic stiffness was different based on age at menopause. Total aortic stiffness (PWV CF ) was lower ( P =0.0019) in late-onset (835±24 cm/s) compared with normal-onset (963±24 cm/s) PMW by a clinically meaningful difference of 128 cm/s ( Figure 1A ). Age at menopause was inversely related to PWV CF (r=−0.27, P =0.0021) ( Figure 1B ). To determine if the age range used for the classification of normal-onset PMW influenced these observations, PWV CF was compared between the late-onset group and a subset of normal-onset PMW restricted only to women who completed menopause at age 50-52 years 46 – PWV CF was 144 cm/s lower ( P =0.0006) in the late-onset PMW compared with this subgroup ( Figure S1A ). Next, the influence of key confounding variables on the relation between age at menopause was assessed using multivariable regression models. A clinically meaningful difference (i.e., >100 cm/s) in aortic stiffness between late- and normal-onset PMW persisted in multiple linear regression models, including the fully adjusted model (−101 cm/s; 95% CI: −188 to −14 cm/s) incorporating age, CVD risk factors and medication use ( Table 2 ). Moreover, the relation between age at menopause expressed as a continuous variable and aortic stiffness remained statistically significant in the fully adjusted model (14 cm/s lower [−27 to −1 cm/s] PWV CF per year of menopause onset) ( Table 2 ). Given the higher BMI and prevalence of obesity in the normal-onset group, a subgroup analysis in PMW with BMI in the “normal” range 40 was performed. PWV CF was 131 cm/s lower ( P =0.010) in late-onset PMW (819±31 cm/s) compared with normal-onset PMW (950±33 cm/s) in this subset with more closely matched BMI (normal-onset PMW: 22.0 kg/m 2 , late-onset PMW: 21.6 kg/m 2 ) ( Figure S1B ). Additionally, to enhance confidence that prior use of hormone therapy did not play a major role in the observations, another sub-analysis was performed in PMW with vs. without a history of hormone therapy use. Late-onset PMW exhibited lower ( P =0.0026, group effect) PWV CF compared with normal-onset PMW independent of prior hormone therapy use, and there was no interaction effect ( P =0.64) between age at menopause and prior hormone therapy use ( Figure S1C ). Statistical adjustment for time since menopause in the whole group did not affect the conclusions related to lower PWV CF in late-onset PMW (adjusted means late-onset PMW: 863±32 cm/s, normal-onset PMW: 947±22 cm/s; P =0.036). The next aim was to determine the influence of age at menopause on load- and structural-dependent stiffness and the contribution of each to age at menopause-associated differences in total aortic stiffness. PWV CF,load was lower ( P =0.0085) in late-onset PMW (−53±8 cm/s) compared with normal-onset PMW (−32±6 cm/s) ( Figure 2A ) and there was a trend for an inverse correlation between PWV CF,load and age at menopause (r=−0.17, P =0.055) ( Figure 2B ). In addition, PWV CF,structural was 107 cm/s lower ( P =0.015) in late-onset PMW (888±23 cm/s) compared with normal-onset PMW (995±25 cm/s) ( Figure 2C ) and inversely correlated with age at menopause (r=−0.21, P =0.016) ( Figure 2D ). Most of the difference in total aortic stiffness (128 cm/s) between late- and normal-onset PMW was a result of differences in structural stiffness as nearly 85% (i.e., 107 cm/s out of 128 cm/s) of the difference in total aortic stiffness between late- and normal-onset PMW was attributable to structural-dependent stiffness, whereas <20% (i.e., 21 cm/s out of 128 cm/s) was a result of load-dependent stiffness. Given the findings that structural-dependent stiffness (rather than load-dependent stiffness) appeared to be the primary driver of age at menopause-related differences in aortic stiffness, subsequent experiments were performed to assess if the circulating milieu is a contributing mechanism of lower structural-dependent stiffness in late-onset menopause by assessing aortic intrinsic stiffness in aortas from mice treated with serum from a subset of late-onset (n=11) and normal-onset (n=13) PMW. PWV CF and PWV CF,structural were lower in late-onset compared with normal-onset PMW in this subset ( Table S2 ). The participant characteristics of this subset were largely reflective of the overall cohort in terms of age and other major cardiometabolic risk factors, including BP, total, HDL and LDL cholesterol, triglycerides and glucose, physical activity level and BMI ( Table S2 ). Serum from late-onset PMW induced lower ( P =0.019) aortic elastic modulus (4905±647 kPa) compared with serum from normal-onset PMW (7334±329 kPa) ( Figure 3A ). Age at menopause was inversely correlated with ex vivo aortic elastic modulus after serum exposure ( Figure 3B ), and aortic elastic modulus was positively correlated with PWV CF and PWV CF,structural in PMW ( Figure 3C and 3D ). Differences in aortic elastic modulus between late- and normal-onset PMW remained when statistically accounting for cardiometabolic risk factors ( Table S3 ). Next, the mitochondrial antioxidant MitoQ was used to determine the contribution of mitoROS to serum-induced stiffening. Co-incubation of aortic rings with MitoQ during serum exposure from late-onset and normal-onset PMW lowered aortic elastic modulus in late-onset PMW (from 5304±677 kPa to 4028±312 kPa with MitoQ; P =0.016) and normal-onset PMW (from 7495±320 kPa to 4912±283 kPa with MitoQ; P =6.9x10 -6 ) (interaction effect: P =0.064) ( Figure 3E ). The absolute difference in elastic modulus after MitoQ incubation was smaller ( P =0.029) in late-onset PMW (1276±349 kPa) compared with normal-onset PMW (2583±391 kPa) ( Figure 3F ). In addition, age at menopause was inversely correlated with the mitoROS-related contribution to aortic elastic modulus in PMW ( Figure 3G ). To identify potential molecular transducers contributing to the effects of the circulating milieu on aortic intrinsic stiffness, untargeted lipidomics analyses (774 lipids) were performed on serum from the subset of PMW used for the serum exposure experiments. Twenty lipid-related metabolites that significantly differed between groups were identified: 18 lipids were lower, and 2 lipids were higher in late-onset PMW compared with normal-onset PMW ( Figure 4A and B ). Ten out of the 20 lipid metabolites that differed between groups were TG-related, which were all lower in late-onset PMW ( Figure 4A and B ). Of the TG-related metabolites that differed between groups, four were significantly associated with the mitoROS-related contribution to elastic modulus after serum exposure from PMW ( Figure 4C ). Of these 4, the greatest difference between late- and normal-onset PMW was observed for TG(16:0_24:0_18:1) (Log 2 FC=−2.02) ( Figure 4B ). To determine if the differences in TG(16:0_24:0_18:1) contributed to the observed differences in aortic intrinsic stiffness based on age at menopause by inducing mitoROS, aortic rings were exposed to a structurally similar TG-related lipid, TG(16:0), and elastic modulus was assessed with or without co-incubation of MitoQ. The concentration of TG(16:0) corresponded to the difference observed between late- and normal-onset PMW of 58μM. Aortic rings exposed to control media + this concentration of TG(16:0) had higher ( P =0.033) elastic modulus (4748±544 kPa) compared with aortic rings exposed to control media without TG(16:0) (3017±317 kPa) ( Figure 4D ). Co-incubation of TG(16:0) with MitoQ lowered ( P =0.030) aortic elastic modulus (2989±544 kPa) to levels observed in control media; however, co-incubation of MitoQ with control media had no effect on aortic elastic modulus ( Figure 4D ).

In the current study, we provide initial evidence that late-onset PMW exhibit lower aortic stiffness compared with normal-onset PMW, which is at least partly independent of traditional CVD and female-specific risk factors. Leveraging recently developed participant-specific modeling approaches 12 , 31 - 33 , we demonstrate that although lower load-dependent stiffening plays a role, lower structural-dependent stiffness is the primary contributor to attenuated total aortic stiffness in late-onset PMW. Using translational ex vivo bioassays 20 , we show the circulating milieu is an important mechanism contributing to lower aortic intrinsic stiffness in late-onset PMW, demonstrate a role of reduced mitoROS-induced oxidative stress in lower circulating milieu-evoked intrinsic stiffness and identify TG-associated lipid metabolites as potential circulating factors that possibly contribute to mitoROS-related effects on aortic intrinsic stiffness based on age at menopause. Collectively, our study provides important mechanism-based evidence that late-onset menopause attenuates aortic stiffness, which may in part explain reduced CVD risk in this group. Age at menopause represents an important female-specific factor that influences CVD risk in women after menopause, as recently recognized by the American Heart Association and the American College of Cardiology 56 , 57 . Large epidemiological studies demonstrate women who complete menopause at a later age exhibit lower CVD risk than women who complete menopause at a normal or earlier age 5 , 7 - 9 , but the physiological mechanisms that explain differences in risk based on age at menopause are not well understood. We recently demonstrated that vascular endothelial function is greater in late-onset compared with normal-onset PMW, which may contribute to the lower CVD risk in late-onset menopause 26 . Aortic stiffening occurs with advancing age and accelerates with the menopause transition 11 , 17 . Here, we demonstrate that compared with normal-onset PMW, late-onset PMW exhibit a near 15% lower total aortic stiffness, as assessed by the reference-standard measure of PWV CF 10 . The difference in PWV CF between late- and normal-onset PMW was ~130 cm/s, which is highly clinically meaningful as 100 cm/s lower PWV CF is associated with a 15% lower risk of CV mortality 58 . We also found a positive, linear relation between age at menopause and aortic stiffness, and differences in aortic stiffness persisted when comparing late-onset PMW to a more restricted group of normal-onset PMW with an age at menopause of 50-52 years, suggesting the effect of age at menopause on aortic stiffness is not dependent on the precise age ranges used to define normal- or late-onset menopause. Although several key traditional risk factors, co-morbidities and use of medications were well-matched between groups, late-onset PMW were generally healthier. Indeed, late-onset PMW had lower BMI, less obesity, lower triglycerides and evidence of an overall reduced burden of other risk factors consistent with the lower risk of several cardiometabolic diseases and longer lifespan previously observed in this population 8 , 9 , 59 - 64 . However, our multiple linear regression analyses suggest that a clinically meaningful difference in PWV CF based on age at menopause persists when statistically adjusting for many key risk factors. Moreover, differences in aortic stiffness were still observed in a subset of our cohort in the “normal” BMI range. Large-elastic artery stiffness represents a primary therapeutic target to reduce the burden of CVDs 10 ; as such, our findings provide initial evidence that age at menopause exerts at least some independent influence on aortic stiffness in the postmenopausal period. The impact of age at menopause on aortic stiffness was also largely independent of other female-specific factors known to modulate vascular aging in PMW 17 , 57 as we did not observe notable differences in terms of prior use of oral contraception, history of natural versus surgical menopause, prior hormone therapy use and circulating sex hormones. In addition, all of the PMW in our analysis had been postmenopausal for at least 5 years (~15 years on average), reducing the possibility that our observations were influenced by the more variable hormone levels associated with the menopause transition 17 , 45 . Although time since menopause was shorter in late-onset PMW because of a similar chronological age between groups, lower aortic stiffness in late-onset PMW was still observed when this difference was accounted for statistically. Lastly, in our cohort of late-onset PMW, a similar age at menarche accompanied by a later age of menopause led to a longer reproductive lifespan, which has been associated with lower CVD risk 65 , 66 . Aortic stiffening is a key risk factor for hypertension as intrinsic stiffening of the aorta with aging is thought to underlie age-dependent elevations in BP, largely by augmenting systolic BP 10 , 67 . Higher BP, in turn, further increases aortic stiffness via greater mechanical loading of the arterial wall 10 . To determine the BP-dependent and -independent mechanisms of aortic stiffness, we employed recently established participant-specific exponential models to separate total aortic stiffness into load- and structural-dependent components 12 , 31 - 33 . We found that load-dependent stiffness was 21 cm/s lower in late- versus normal-onset PMW. Although this observation is not entirely surprising given the lower systolic and diastolic BP observed in late-onset PMW, the finding is important because load-dependent aortic stiffening is independently associated with CVD and end organ damage 31 , 33 , 34 . However, most of the difference in total aortic stiffness between late- and normal-onset PMW was related to differences in structural stiffness, which was 107 cm/s lower in late-onset PMW. This observation is clinically relevant as greater structure-dependent stiffness is an independent predictor of mortality and adverse CV events 31 , 34 . These findings suggest that the intrinsic stiffness of the aorta is an important factor in the observed differences in aortic stiffness based on age at menopause. A mechanism of increased intrinsic stiffness of the arterial wall with aging and menopause is unfavorable changes to the circulating milieu – the collection of bioactive factors in the bloodstream 15 . We have shown that serum from older men and PMW increases intrinsic stiffness of aortas from young adult mice ex vivo compared with serum from younger men and premenopausal women 20 . Here, we extend these findings to show that the serum-evoked increase in aortic intrinsic stiffness is lower with serum from late-onset PMW in comparison with serum from normal-onset PMW. These observations persisted after adjustment for CV-related variables, demonstrating that the circulating milieu has an independent contribution to age at menopause-associated differences in aortic intrinsic stiffness. The mitochondria-targeted antioxidant MitoQ ameliorated increases in aortic intrinsic stiffness in both normal-onset and late-onset PMW, consistent with the idea that excessive mitoROS contributes to circulating milieu-dependent stiffening of the aorta. This observation is in line with our findings that chronic supplementation with MitoQ reduces aortic stiffness in older adults 23 , 24 . Importantly, the contribution of mitoROS to aortic intrinsic stiffness was smaller in late- compared with normal-onset PMW, indicating that less mitoROS partially underlies lower intrinsic stiffness in late-onset menopause. Persistent excess mitoROS increases aortic intrinsic stiffness by promoting elastin degradation and collagen deposition 22 ; however, changes in these structural proteins are unlikely given the short duration of serum exposure in our ex vivo experiments. Oxidative stress can also induce aortic stiffening more rapidly by increasing endothelial and vascular smooth muscle cell stiffness 15 , 68 and/or increasing stiffness-promoting advanced glycation end products 69 . As such, it is possible that lower vascular cell stiffness and/or advanced glycation end products secondary to reduced mitoROS-associated oxidative stress may contribute to reduced aortic intrinsic stiffness in late-onset PMW. We have previously shown that the circulating lipid milieu differs between late- and normal-onset PMW and differences in select lipid-related metabolites induce endothelial dysfunction by promoting mitoROS 26 , 70 . However, whether circulating lipidome-associated modulation of mitoROS is in part responsible for differences in aortic stiffness in relation to age at menopause is not known. Thus, we performed a new lipidomics analysis and identified 20 lipids that differed between late- and normal-onset PMW, most of which were TG-related lipid species. Several TG-related lipids were negatively associated with mitoROS-dependent aortic elastic modulus. TG(16:0_24:0_18:1) had the highest fold change difference between groups and is the TG-related metabolite that we have previously shown contributes the effects of the circulating milieu on endothelial cell mitoROS 26 . Incubation of aortic rings with the difference in TG(16:0) observed between groups induced aortic stiffness by increasing mitoROS, implicating lower levels of TG(16:0) as a probable mechanism of lower mitoROS-related intrinsic aortic stiffness in late-onset PMW. There are several limitations to our study that should be acknowledged. Our cohort lacked racial and ethnic diversity, limiting the generalizability of our findings. We also acknowledge that unmeasured and/or insufficiently captured factors may have contributed to the observed differences attributed to age at menopause, including undiagnosed or unreported clinical disease or differences in specific forms or duration of use of prior hormone therapy, resulting in residual confounding. In addition, the directionality of the relation between age at menopause and arterial function could not be determined and reverse causality is a possibility. Indeed, age at menopause is linked with ovarian follicle count 71 , 72 , which is likely influenced by ovarian arterial blood flow 73 . As such, it is possible that better vascular function may contribute to delaying the onset of menopause in women. There are also a few limitations to our ex vivo serum exposure experiments including the small sample size, likely inability of the assay to capture changes in collagen deposition and/or elastin degradation – processes that are thought to occur over years 10 , 74 – and the fact that we are the first group to use the approach to assess mechanisms contributing to phenotypic differences in aortic stiffness. As such, replication and extension of our findings by other groups in larger cohorts are essential to enhance confidence in our results and gain additional insight into mechanisms. Despite these limitations, the clinical relevance of aortic intrinsic stiffness determined with this approach is supported by our observations of positive linear relations between ex vivo aortic intrinsic stiffness and the in vivo measures of aortic stiffness in the subjects from whom the serum was collected. Lastly, PMW who complete menopause before age 45 (premature or earlier menopause) have higher CVD risk 17 , 75 , 76 , but if these women exhibit accelerated aortic stiffening has not been investigated and represents an important research gap in women’s CV health.

Cardiovascular disease (CVD) remains the leading cause of mortality among women worldwide 1 . Although CVD risk increases with advancing age and is higher after menopause in women 2 , 3 , age at menopause is an emerging female-specific factor that modulates CVD risk in the postmenopausal period 3 , 4 . Indeed, most women complete menopause between the ages of 45 and 54 years (normal-onset), but ~10% of women complete menopause at age 55 or later (late-onset) 5 , 6 . Large cohort studies demonstrate that late-onset postmenopausal women (PMW) have 10-20% lower risk of CVD compared with normal-onset PMW 5 , 7 - 9 , but the complete physiological mechanisms contributing to reduced CVD risk are not fully understood. Increased aortic stiffness, as assessed by the reference-standard measure of carotid-femoral pulse wave velocity (PWV CF ) 10 , is an independent predictor of CV events and mortality 11 . A primary mechanism of aortic stiffening with aging is an increase in the intrinsic stiffness of the aorta, i.e., augmented structural-dependent aortic stiffening 10 , 12 . Structure-dependent stiffness is largely determined by structural proteins and their cross linking by advanced glycation end products in the arterial wall as well as the inherent stiffness of vascular cells in the aorta 10 , 13 - 15 . Increases in structural-dependent stiffness contribute to age-related elevations in blood pressure (BP), which, in turn, increases load-dependent aortic stiffening, i.e., the stiffness of the aorta attributable to the mechanical distending pressure placed on the arterial wall 10 , 16 . Although aortic stiffness is higher in PMW in comparison with premenopausal women 17 - 19 , whether late-onset compared with normal-onset PMW have lower aortic stiffness, and the relative contributions of structural- and load-dependent stiffening, has not been investigated. The circulating milieu, the collection of bioactive factors in the bloodstream that can induce physiological effects on tissues and cells, is an emerging mechanism of aortic intrinsic (structural) stiffness with aging and menopause 20 , 21 . We have shown that aortic intrinsic stiffness, assessed by elastic modulus in isolated aortas from young mice, is higher when exposed to serum from older men and PMW compared with serum from young men and premenopausal women 20 . A primary mechanism by which the circulating milieu induces adverse effects in the vasculature is excessive mitochondria-derived reactive oxygen species (mitoROS)-related oxidative stress 22 - 25 . Lipid-related metabolites are key circulating molecular transducers that augment mitoROS-related vascular oxidative stress 26 - 28 and many of these species have been linked to aortic stiffening with aging and menopause 29 , 30 . However, whether the circulating milieu of late-onset PMW is a mechanism contributing to lower aortic intrinsic stiffness and if this is in part because of a modulation of mitoROS-induced oxidative stress and bioactive lipids is unknown. We recently demonstrated that compared with normal-onset PMW, late-onset PMW have preserved vascular endothelial function because of lower mitoROS-dependent oxidative stress and the circulating lipid milieu 26 . The purpose of the current study was to extend these findings and assess the influence of late-onset menopause on aortic stiffness and the contributing mechanisms. We first aimed to determine if late-onset PMW had lower total aortic stiffness compared with normal-onset PMW. We then assessed load- and structure-dependent aortic stiffness using a novel mathematical model-based approach 12 , 31 - 34 and determined the contribution of each to differences in total aortic stiffness between late- and normal-onset PMW. Next, we used a reverse translation-based approach to gain insight into the potential role of the circulating milieu in transducing some of the effects of age at menopause on aortic stiffness by measuring aortic intrinsic stiffness in aortic rings after exposure of serum from PMW. Lastly, we determined the mechanistic role of circulating milieu-evoked increases in mitoROS in age at menopause-related differences in aortic intrinsic stiffness and identified circulating lipid metabolites that may contribute.

In summary, we provide evidence demonstrating that PMW who complete menopause at a later age have lower aortic stiffening compared with PMW who complete menopause at a normal age. Reduced aortic stiffness in late-onset PMW is in part because of lower load-dependent aortic stiffening but primarily a result of attenuated structure-dependent stiffening. Lower structural-dependent stiffening in late-onset PMW is linked with a more favorable composition of circulating lipid metabolites and a suppression of mitoROS-related oxidative stress. These observations extend our previous work 26 and provide evidence that age at menopause is an important female-specific factor that may influence CVD risk in part by modulating aortic stiffening. Our findings support the need for clinical attention to age at menopause-associated differences in aortic stiffness in preventative and individualized care aimed at reducing CVD risk in women after menopause.