Sex Differences in the Relation Between Frailty and Endothelial Dysfunction in Old Mice

Abstract

Vascular endothelial function declines with age on average, but there is high variability in the magnitude of this decline within populations. Measurements of frailty, known as frailty index (FI), can be used as surrogates for biological age, but it is unknown if frailty relates to the age-related decline in vascular function. To examine this relation, we studied young (4–9 months) and old (23–32 months) C57BL6 mice of both sexes. We found that FI was greater in old compared with young mice, but did not differ between old male and female mice. Middle cerebral artery (MCA) and mesenteric artery endothelium-dependent dilation (EDD) also did not differ between old male and female mice; however, there were sex differences in the relations between FI and EDD. For the MCA, FI was inversely related to EDD among old female mice, but not old male mice. In contrast, for the mesenteric artery, FI was inversely related to EDD among old male mice, but not old female mice. A higher FI was related to a greater improvement in EDD with the superoxide scavenger 4-hydroxy-2,2,6,6-tetramethylpiperidin-1-oxyl in the MCAs for old female mice and in the mesenteric arteries for old male mice. FI related to mesenteric artery gene expression negatively for extracellular superoxide dismutase (Sod3) and positively for interleukin-1β (Il1b). In summary, we found that the relation between frailty and endothelial function is dependent on sex and the artery examined. Arterial oxidative stress and proinflammatory signaling are potential mediators of the relations of frailty and endothelial function.

With advancing age, there is typically an increase in variability among the population for measures of physiologic function (1). This variability arises as individuals do not all age at similar rates, and chronological age does not necessarily reflect biological age (2). Greater variability in older compared with younger groups is found for measures of vascular function, and may explain the variability in cardiovascular, cerebrovascular, and mortality risk among populations (2). Among human populations, the effects of environment and genetics are often cited as causes of this increased variability with aging (3). However, there is high variability in vascular function among old rodents, despite the high genetic similarity (littermates, inbred strains) and controlled housing environment (4). Thus, understanding the contributors to the variability in vascular function among old rodents will provide insight into the underlying biological causes of vascular aging.

Measures of frailty have been used to understand the heterogeneity among older humans or rodents (2). According to Kane et al., frailty is the “high vulnerability to adverse health outcomes which reduces the capacity to react to stressors” (5). Noninvasive measurements of frailty, known as a frailty index (FI), are used as surrogates for biological age and are predictive of mortality, cardiovascular diseases, and dementia (2,6–8). The use of a FI is now encouraged to evaluate the overall health across several domains during interventions (9). Mouse frailty can be assessed using a 31-item clinical FI based on established clinical signs of deterioration (10). Models based on this mouse FI were created to predict chronological age (Frailty Inferred Geriatric Health Timeline [FRIGHT]) and life expectancy (Analysis of Frailty and Death [AFRAID]) (11). As an estimate of biological age, FI could be mechanistically linked to the variability among older populations; however, the relation of frailty and vascular function is unclear. In these studies, we sought to determine the relation between FI and vascular function across the life span in mice.

A key feature of vascular dysfunction with advancing age is an impairment of endothelial cell function (12). Age-related impairments to cerebral artery endothelial function have detrimental consequences to brain function, and possibly contribute to neurodegenerative diseases (13). Mesenteric artery endothelial function is also impaired with aging (14) and plays a pivotal role in blood pressure regulation (15). The primary cause of age-related endothelial dysfunction is an increase in oxidative stress and inflammatory signaling (12). However, the associations between frailty and vascular endothelial dysfunction, oxidative stress, or inflammatory signaling are unknown.

Sex differences exist for both the rate of decline in vascular function and the magnitude of increase in frailty with advancing age (16,17). In humans, older females are typically more frail, yet the relation between frailty and mortality is stronger in males (16). There are also sex differences in the relations of frailty with age-related cardiac remodeling and with inflammatory cytokines (18,19). Thus, it is important to consider sex differences when examining the relation of frailty to vascular function.

We hypothesized that greater FI is inversely related to endothelial function, measured by ex vivo endothelium-dependent dilation (EDD), in middle cerebral arteries (MCAs) and mesenteric arteries among old mice. We explored this hypothesis for a combined group, as well as groups divided by sex. We further hypothesized that oxidative stress-mediated suppression of EDD, as well as greater prooxidant and proinflammatory gene expression, are related to FI. We also hypothesized that FRIGHT and AFRAID would also be related to EDD. To test these hypotheses, we measured FI, MCA and mesenteric artery EDD, and markers of oxidative stress and inflammatory signaling in male and female C57BL/6 mice ranging in age from 4 to 32 months.

Method

Animals

Male and female C57BL/6 mice were studied. Old mice were obtained from the National Institute on Aging (NIA) colony at Charles River and young mice were purchased from Charles River. Young mice were 4–9 months of age (n = 44) and old mice were 23–32 months of age (n = 49). The main set of the old mice were from the same birth cohort of the NA aging colony. Mice from this cohort were studied at n = 5/sex at ~23, ~24.5, ~26, and ~28 months, although some mice died of natural causes before their study date. All mice were on a normal chow diet (PicoLab Rodent Diet 20, #5053 LabDiet, St. Louis, MO) with ad libitum food and water and were housed in an animal care facility on a 12/12-hour light–dark cycle at 24 °C. Mice were euthanized by exsanguination under inhaled isoflurane prior to vascular reactivity studies. All animal procedures conform to the Guide to the Care and Use of Laboratory Animals (eighth edition, revised 2011) and were approved by the Institutional Animal Care and Use Committee at the University of Oregon.

Mouse Frailty Assessments

Mouse frailty was assessed using a previously studied 31-item FI based on established clinical signs of deterioration in C57BL/6J mice (10). See Supplementary Methods for additional details.

Vascular Reactivity

EDD was assessed ex vivo in isolated, pressurized mesenteric arteries and MCAs as previously described in detail (20). All arteries were submaximally preconstricted with phenylephrine and increases in luminal diameter in response to increasing concentrations of the endothelium-dependent dilator acetylcholine (ACh) or endothelium-independent dilator sodium nitroprusside (SNP) were determined. To determine the superoxide-mediated suppression of EDD, responses to ACh were measured following a 60-minute incubation in the presence of the superoxide scavenger 4-hydroxy-2,2,6,6-tetramethylpiperidin-1-oxyl (TEMPOL) (20). See Supplementary Methods for additional details.

Gene Expression

In samples of mesenteric arteries, mRNA gene expression was quantified for proinflammatory interleukin-1β (Il1b), prooxidant enzyme NADPH Oxidase 2 (Nox2), and antioxidant enzymes superoxide dismutases 1, 2, and 3 (Sod1, Sod2, and Sod3, respectively). mRNA expression was calculated using the 2^−ΔΔCT method. 18S rRNA was used as a housekeeping gene control, and values were normalized to the young group. See Supplementary Table 1 for primer sequences. See Supplementary Methods for additional details.

Statistical Analysis

Statistical analyses were performed with IBM SPSS (version 24, Armonk, NY). Group differences were assessed by 2 × 2 (Age × Sex) analysis of variance (ANOVA), and in cases of a significant F value, post-hoc analyses were performed with a Bonferroni correction for preplanned comparisons. For cases where only 2 groups were compared, a t test for independent samples was used. Levene’s test was used to determine differences in variability among groups. For dose responses, group differences were determined by repeated measures ANOVA. Pearson correlation analysis was used to assess bivariate relations of interest. Partial correlations were used to assess the influence of age or baseline EDD. Significance was set at p <.05. Values are presented as mean ± SEM. See Supplementary Methods for additional details.

Results

Animal Characteristics and Frailty

Body mass was greater for old mice compared with young mice, and for male mice compared with female mice (p < .05; Supplementary Table 2). There were several age, sex, or Age × Sex interaction effects for tissue mass, as shown in Supplementary Table 2. FI was greater in old mice (old female: 0.18 ± 0.02, old male: 0.19 ± 0.01) compared with young mice (young female: 0.02 ± 0.00, young male: 0.05 ± 0.00, ANOVA main effect of age p < .05). When we further split the old group, we found that males aged 22–23 months had a lower FI than males aged 26–28 (Supplementary Figure 1B). However, for females, mice aged 23–24 months had a similar FI to those aged 26–28 months (Supplementary Figure 1A). This suggests that frailty continues to increase across the old age group in males, but not in females in this study cohort. We also compared the individual components of FI between only old male and female mice. The old female mice, compared with old male, had a higher mean score and/or frequency for parameters in the ocular/nasal category (cataracts, corneal opacity), alopecia, and rectal prolapse. The old male mice, compared with old female, had a higher mean score and/or frequency for parameters in the discomfort category (grimace scale, piloerection), the physical/musculoskeletal category (distended abdomen, kyphosis, tail stiffening, gait disorders, tremor), dermatitis, and penile/vaginal prolapse (Supplementary Tables 3 and 4).

Impaired Endothelial Function With Age

Mesenteric artery EDD, measured as the vasodilation in response to ACh, was 16% lower in old mice compared with young mice (Figure 1A; maximal response: p = .003, dose response: p = .04). For the MCA, EDD was 25% lower in old compared with young mice (Figure 1B; maximal response: p = .005, dose response: p = .006). Within the group of old mice, there was significantly more variability in mesenteric artery and MCA maximal EDD when compared with the variability within the young mice (Figure 1C and D; p < .05). There was no difference in endothelium-independent dilation (measured as the vasodilation in response to SNP), maximal artery diameters, or preconstriction between young and old mice (Figure 1E and F; Supplementary Table 5; p > .05). These results indicate that both mesenteric artery and MCA endothelial function are impaired with age, but there is a high amount of variability in these age-related impairments.

Figure 1.

Impaired endothelium-dependent dilation in old, compared with young, C57BL/6 mice. (A) Mesenteric artery and (B) middle cerebral artery (MCA) dose response to endothelium-dependent dilator acetylcholine (ACh) in young and old mice. (C) Mesenteric artery and (D) MCA maximal responses to ACh in groups divided by age and sex. (E) Mesenteric artery and (F) MCA dose response to endothelium-independent dilator sodium nitroprusside in young and old mice. *p < .05 vs young. ^†p < .05 vs young female. ^‡p < .05 vs young male. N = 7–39 per group. Values are mean ± SEM.

The effect of age on EDD was consistent for both sexes, as mesenteric artery and MCA maximal EDD were lower in old compared with young within each sex (Figure 1C and D; p < .05). Furthermore, there was no difference between male and female mice at either young or old ages for mesenteric artery or MCA maximal EDD (Figure 1C and D; p > .05). When we further divided the old groups by age, we found that males aged 23–24 months had a higher maximal mesenteric EDD compared with males aged 26–28 and 30–32 months (Supplementary Figure 1D; p < .05). For the MCA, males aged 26–28 months had a higher maximal EDD compared with males 30–32 months (Supplementary Figure 1F; p < .05). These results suggest a continued decline in endothelial function during old age in male mice. In contrast, there were no differences between the old age groups for maximal mesenteric or MCA EDD in female mice (Supplementary Figure 1C and E; p > .05); however, this conclusion is limited by the lack of female mice over the age of 28 months in this study.

Relation of Frailty and Endothelial Function

When all mice (young and old) are analyzed as a group, mesenteric artery and MCA maximal EDD were inversely related to FI (Supplementary Figure 2A and B; p < .01). As chronological age is assumed to be a major driver of these relations, we conducted further analyses with just the old mice. Furthermore, as male and female mice may age differently, we divided the analysis by sex. For the mesenteric artery, there was a significant inverse relation between FI and maximal EDD among old male mice (Figure 2A), but among old females this relation was not significant and trended in the opposite direction (Figure 2B). When controlling for age, the relation between FI and maximal mesenteric artery EDD remained significant for old male mice (r = −.33, p = .045, partial correlation). For the MCA, there was a significant inverse relation between FI and maximal EDD among old female mice (Figure 2D), but this relation was not significant among old male mice (Figure 2C). When controlling for age, the relation between FI and maximal MCA EDD remained significant for old female mice (r = −0.55, p = .011, partial correlation). These results indicate that frailty is related to endothelial function independent of age, but this is not consistent between the sexes or vascular beds.

Figure 2.

Greater frailty index relates to impaired endothelium-dependent dilation in old mice. In old (23–32 months) C57BL6 (A) male and (B) female mice, relation of frailty index to mesenteric artery maximal endothelium-dependent dilation to acetylcholine (ACh). In old C57BL6 (C) male and (D) female mice, relation of frailty index to middle cerebral artery (MCA) maximal endothelium-dependent dilation to ACh.

Oxidative Stress and Inflammatory Signaling

To assess the magnitude of suppression of EDD due to oxidative stress, we repeated the measures of EDD after ex vivo incubation with the superoxide scavenger TEMPOL. For mesenteric arteries from young mice, TEMPOL did not change the response to ACh (Figure 3A; p > .05). For mesenteric arteries from old mice, TEMPOL improved the dose response to ACh and the sensitivity to ACh (half maximal effective concentration [EC₅₀], Figure 3B; Supplementary Table 5; p < .05), but maximal EDD did not change with TEMPOL (Figure 3B; p > .05). For MCAs from both young and old mice, TEMPOL did not affect the maximal response, dose response, or sensitivity to ACh (Figure 3C and D; Supplementary Table 5; p > .05). Thus, in this group of old mice, scavenging superoxide improved sensitivity to ACh in mesenteric arteries, but did not improve maximal EDD in either artery.

Figure 3.

Greater frailty index relates to superoxide-mediated suppression of endothelium-dependent dilation in the mesenteric artery for old male mice and middle cerebral artery (MCA) for old female mice. In (A) young and (B) old mice, mesenteric artery dose response to endothelium-dependent dilator acetylcholine (ACh) in the absence or presence of superoxide scavenger 4-hydroxy-2,2,6,6-tetramethylpiperidin-1-oxyl (TEMPOL) (n = 27–47 per group). In (C) young and (D) old mice, MCA dose response to ACh in the absence or presence of superoxide scavenger TEMPOL (n = 22–48 per group). In old (E) male and (F) female mice, relation of frailty index to the change in mesenteric artery maximal ACh response in the presence of TEMPOL. In old (G) male and (H) female mice, relation of frailty index to the change in MCA maximal ACh response in the presence of TEMPOL. Values are mean ± SEM.

We further sought to determine if the suppression of EDD by superoxide was related to frailty. Among mesenteric arteries from old male mice, suppression of EDD by superoxide (the difference of the maximal ACh response and maximal ACh response in the presence of TEMPOL) was related to FI (Figure 3E). In contrast, among old female mice, the inverse relation between FI and suppression of mesenteric artery EDD by superoxide was found (Figure 3F). For the MCAs, suppression of EDD by superoxide was related to FI among old female mice (Figure 3H), but not related in old male mice (Figure 3G). These results suggest that frailty is related to oxidative stress in arteries among old mice; however, these relations are dependent on sex and vascular bed. These results for oxidative stress-mediated suppression follow the same trends as those for maximal endothelial function (ie, significant in mesenteric arteries for males and in MCAs for females). It is important to note that the relations of FI and suppression of EDD by superoxide are no longer significant when controlling for baseline EDD (p > .05, partial correlations). Thus, in general, those mice with a lower endothelial function had an improvement when superoxide was scavenged, whereas those with a higher endothelial function had a reduction when superoxide was scavenged.

To further investigate the relations of frailty with oxidative stress and inflammatory signaling, we measured gene expression in the mesenteric arteries. We found that gene expression of antioxidants Sod2 and Sod3 was lower in old compared to young mesenteric arteries (Figure 4B and C; p < .05), while Sod1 and prooxidant enzyme Nox2 did not differ between groups (Figure 4A and D; p > .05). Gene expression of proinflammatory cytokine Il1b was greater in old male compared to young male mesenteric arteries (p = .01), but did not differ between age groups for female mice (Figure 4F; p = .28). Mesenteric artery Il1b gene expression was also greater in males compared with females when compared at both young and old ages (p < .05), but there were no sex differences for any of the other genes (Figure 4A–E; p > .05). When all mice are included, FI was negatively related to gene expression of Sod3 (p = .001) and positively related to gene expression of Il1b (p = .04); however these relations were no longer significant when controlling for age (partial correlation, p > .05). FI did not relate to Sod1, Sod2, or Nox2 gene expression (p > .05). These results indicate that frailty is related to decreased antioxidant and increased proinflammatory gene expression in arteries, but chronological age may be the underlying mediator of these relations.

Figure 4.

Differences in mesenteric artery gene expression with age and relation to frailty index. Gene expression for antioxidants (A) superoxide dismutase 1 (Sod1), (B) Sod2, and (C) Sod3, (D) prooxidant NADPH oxidase 2 (Nox2), and (E) proinflammatory interleukin-1β (Il1b) in young and old mice (n = 6–22 per group). Relation of frailty index to (F) Sod3 and (G) Il1b in a combined group of mice (n = 53). *p < .05 vs young female. ^†p < .05 vs young male. ^‡p < .05 vs old female. Values are mean ± SEM.

FRIGHT and AFRAID Indices

Using the individual components of FI, we calculated FRIGHT age and AFRAID score (11), and these did not differ by sex (Supplementary Table 2). In addition, FRIGHT age and AFRAID score were not related to mesenteric artery or MCA maximal EDD for either old males or old females (p > .05, data not shown). Thus, in contrast to FI, we find no relation between endothelial function and FRIGHT age or AFRAID score.

Discussion

In this study, we found that despite a very high genetic and environmental similarity, there was a large amount of variability in vascular function among old mice. We further found that among old mice, frailty is related to vascular function, but this is dependent on the sex of the mouse and the vascular bed studied. Importantly, these relations of frailty and endothelial function were significant independent of age, suggesting that “biological age” may be an important factor in the variability of vascular function among old mice. We further found that arterial oxidative stress and inflammatory signaling are related to frailty, indicating that these may have a mechanistic role in the relation of vascular function and frailty. Figure 5 presents a summary of these findings.

Figure 5.

Summary of findings. Figure created with Biorender.com.

Relation of Frailty and Vascular Function

FI has become an essential assay in aging research and is now encouraged in studies of interventions targeting life span and healthspan (9). Yet, the data on the relation of frailty and vascular function are limited. A recent attempt at a meta-analysis of frailty and vascular outcomes was unsuccessful due to a lack of consistency in techniques to measure both frailty and vascular function in humans (21). One previous study found a relation of flow-mediated dilation and frailty among patients with chronic kidney disease (22). Another study found a relation of frailty and asymmetric dimethylarginine, an inhibitor of nitric oxide synthases, in a large sample from the Toledo Study for Healthy Aging (23). As the results of the present study are correlational, it remains unknown what factor is driving the relations. Potentially, age-related vascular impairments cause increased frailty, such as by leading to reduced organ blood flow. On the other hand, frailty may serve as a marker of the underlying biological age, the mechanisms of which are potential mediators of age-related vascular impairment.

In this study, we used the 31-item clinical FI created for mice by Whitehead et al. (10). We chose this scale as it does not require specialized equipment, unlike other available measures of frailty in mice (24). However, as other measures of frailty include functional outcomes (blood pressure, physical activity, etc.), these could be better related to endothelial function than the clinical FI. We also calculated the FRIGHT age and AFRAID score, measures created using machine learning to predict chronological age and death using individual items of the clinical FI (11). We found no relation between vascular function and FRIGHT age or AFRAID score in this study; however, a limitation is that we did not measure body weight change from 21 months of age (a component of FRIGHT/AFRAID, but not FI) or vision loss. Thus, while we found relations of the 31-item FI with vascular endothelial dysfunction, future studies should explore if other measures of frailty relate to vascular function.

Oxidative Stress and Inflammatory Signaling as Mediators of Frailty

Increased oxidative stress and inflammatory signaling are known mediators of age-related diseases and dysfunction (12), and thus it is not surprising that these are also related to frailty. Numerous studies in human subjects have related circulating markers of oxidative stress and inflammatory cytokines to measures of frailty (25–27). In our study, we find that frailty relates to local vascular expression of antioxidant and inflammatory genes. We also find that oxidative stress-mediated suppression of endothelial function is related to frailty. It is notable that the superoxide scavenger TEMPOL did not improve maximal EDD in arteries from old mice (on average), in contrast to previous studies (20,28). This illustrates the dual actions of superoxide, both to impair EDD by scavenging nitric oxide, but also as a potential contributor to EDD by conversion to vasodilatory hydrogen peroxide. This is supported by our previous finding that TEMPOL impairs cerebral artery endothelial function in young mice (28). However, we did not measure superoxide in these studies; and thus, we cannot definitively conclude that the actions of TEMPOL were to scavenge superoxide. In summary, our results suggest that oxidative stress is potentially a strong mediator of the relation between frailty and vascular function, but further studies are needed.

Sex Differences in Frailty

In humans and mice, females typically have higher frailty compared with males (10,16,18,29), although some studies find no sex differences for frailty (30,31), including the present study. There are a few potential reasons why we did not find sex differences in contrast to previous studies in mice, including:

1. Some previous studies used retired breeders (18,32) and breeding likely places more stress on the bodies of females, leading to higher frailty.

2. Differences in mouse strain are known to influence FI (30,31), and could account for sex differences. Most previous studies use C57BL/6J mice, while this study used C57BL/6 mice from the NIA rodent colony.

3. There is inconsistency in the age of mice labeled as “old” among studies, and this may contribute to sex differences as female mice develop frailty at a faster rate than males (32).

Thus, findings of sex differences for frailty are likely highly dependent on the population studied. FI scores are proposed to predict biological age, mortality, and the success of interventions; however, sex differences exist for a number of these outcomes. Studies in humans find that males have higher mortality rate for a given FI score compared with females (16,33). FI also relates to cellular cardiac remodeling in male, but not female, mice (19). In summary, sex differences exist not only for measures of frailty, but also for how this frailty relates to mortality and physiological function.

Sex Differences in the Relation of Frailty to Vascular Aging

Historically, most vascular aging studies conducted in rodents used only males, and this has led to a paucity of data about the sex differences in the mechanisms of vascular impairment with advancing age. Our findings of no sex differences in cerebral artery endothelial dysfunction in old mice aligns with previous studies (34,35), although the age of onset for vascular impairment may vary by sex (35), a possibility that we cannot address due to a lack of middle-aged mice in our study. Although there were no sex differences for age-related endothelial dysfunction on average, we did find important differences between females and males for the relation between frailty and endothelial dysfunction. As this is a correlational study, we cannot definitively identify the cause of the sex differences in these relations; however, possible reasons include sex differences in frailty parameters, inflammation, and circulating sex hormones.

As the clinical FI is a composite score of multiple body systems, it is possible for 2 groups to have similar overall scores, but to attain these scores for different reasons. In our study, we find that old females, compared with old males, had higher scores for and/or more frequent alopecia and items in the ocular/nasal category. Old males, compared with old females, had higher scores for and/or more frequent dermatitis and items in the physical/musculoskeletal and discomfort categories. Some of these results align with previous findings, for example higher scores for alopecia are consistent in females, and higher scores for piloerection and distended abdomen are consistent for males (18). In male mice, frailty is related to mesenteric artery dysfunction. As the mesenteric arteries contribute to blood pressure regulation and perfusion of the gut, it seems reasonable that the function of these arteries is associated with frailty that manifests as signs of discomfort and general poor health (distended abdomen, gait disorders). Old male mice treated with enalapril, an angiotensin converting enzyme inhibitor, have lower blood pressure and frailty, specifically related to lower scores for items in the physical/musculoskeletal category (32), supporting the connection between blood pressure and these frailty parameters. In females, the relation of cerebral artery impairment and poor ocular frailty scores is interesting given that outcomes in the eye are proposed as Alzheimer’s disease biomarkers (36). Potentially, the proximity of the eyes and the brain, or the common conduit circulation, is leading to impairments in both organs primarily in females. Thus, differences in the parameters are contributing to higher FI scores in males versus females, and how these relate to physiological outcomes, deserves more investigation.

Frailty is related to increased inflammatory cytokines, but there are sex differences for these relations. In humans, C-reactive protein and fibrinogen relate to frailty in females, but not males (37). In mice, circulating IL-6, IL-9, and interferon-γ relate to frailty in females, while circulating IL-12p40 relates to frailty in males (18). It is possible that these cytokines have differential effects on cerebral versus mesenteric circulations, and thus differences in inflammatory signaling could underlie the sex differences in the relation of frailty to endothelial function, a possibility that needs further exploration.

Declining sex hormones may also explain the differences in the relation between frailty and endothelial function between females and males. Estrogen deficiency is associated with both vascular impairments and increased frailty (17,38,39). In contrast in male mice, the increases in frailty and association to mesenteric artery dysfunction may be caused by declining testosterone concentrations (17,40,41). Thus, age-related declines in estrogen in females and testosterone in males are both associated with increased frailty, but different actions of these hormones may underlie the sex differences in the relations of frailty to vascular impairment.

Limitations

Several limitations were unavoidable in this study. Only gene expression measurement was possible because of the small size of tissue that can be collected for mouse mesenteric arteries; and thus, no conclusions can be made about protein concentration or enzyme activity. We also did not directly measure superoxide production in these arteries. While this study does represent a wider range of ages than most previous studies, our cohort does have a gap in the middle ages. Additionally, our findings could be influenced by a survival bias. For example, those mice that survived to 30–32 months were likely healthier in early old age compared to those who did not survive. Lastly, sex differences may differ by mouse strain; and thus, caution should be used when generalizing these results to other strains.

Conclusion

In summary, we found no sex differences for composite FI scores or endothelial impairment among old mice. However, there are sex differences in the individual components of FI and in the relations of frailty to endothelial function, and these may be mediated by sex differences in arterial oxidative stress and proinflammatory signaling. These results suggest that interventions to improve mesenteric artery function in male mice and cerebral artery function in female mice may have benefits for diminishing frailty.

These results highlight that (i) there is a high variability in measures of vascular function among old subjects, even rodents, and these are often ignored by studying mean differences between groups, and (ii) age-related changes in frailty and vascular function are sex dependent and future research needs to direct attention to the potential interactions of age and sex.

Funding

This work was supported in part by awards from the National Institutes of Health AG046326 and AG064016 to A.E.W. and University of Oregon UROP mini-grant to J.A.C.

Conflict of Interest

None declared.