Maladaptive Changes Associated With Cardiac Aging Are Sex-Specific and Graded by Frailty and Inflammation in C57BL/6 Mice

Abstract

We investigated whether late-life changes in cardiac structure and function were related to high levels of frailty and inflammation in male and female mice. Frailty (frailty index), ventricular structure/function (echocardiography), and serum cytokines (multiplex immunoassay) were measured in 16- and 23-month-old mice. Left ventricular (LV) mass and septal wall thickness increased with age in both sexes. Ejection fraction increased with age in males (60.4 ± 1.4 vs 68.9 ± 1.8%; p < .05) but not females (58.8 ± 2.5 vs 62.6 ± 2.4%). E/A ratios declined with age in males (1.6 ± 0.1 vs 1.3 ± 0.1; p < .05) but not females (1.4 ± 0.1 vs 1.3 ± 0.1) and this was accompanied by increased ventricular collagen levels in males. These changes in ejection fraction (r = 0.52; p = .01), septal wall thickness (r = 0.59; p = .002), E/A ratios (r = -0.49; p = .04), and fibrosis (r = 0.82; p = .002) were closely graded by frailty scores in males. Only septal wall thickness and LV mass increased with frailty in females. Serum cytokines changed modestly with age in both sexes. Nonetheless, in males, E/A ratios, LV mass, LV posterior wall thickness, and septal wall thickness increased as serum cytokines increased (eg, IL-6, IL-3, IL-1α, IL-1β, tumor necrosis factor-α, eotaxin, and macrophage inflammatory protein-1α), while ejection fraction declined with increasing IL-3 and granulocyte-macrophage colony stimulating factor. Cardiac outcomes were not correlated with inflammatory cytokines in females. Thus, changes in cardiac structure and function in late life are closely graded by both frailty and markers of inflammation, but this occurs primarily in males. This suggests poor overall health and inflammation drive maladaptive changes in older male hearts, while older females may be resistant to these adverse effects of frailty.

Keywords: Chemokines; Echocardiography; Frailty index; Pro-inflammatory cytokines; Sex differences.

Figure 1.

Age-associated changes in systolic and diastolic function are more prominent in male hearts and are graded by the degree of frailty in males but not in females. (A) Representative M-mode echocardiography recording from an old male mouse. Parameters of interest measured in this example are illustrated. (B) Example of a pulse-wave Doppler recording from an old male mouse. The velocities of mitral valve blood flow, both early (E wave) and later (A wave) are indicated. (C) There were no effects of either age or sex on heart rate (HR) measured with echocardiography in this study. (D, E) There was no relationship between HR and frailty index (FI) scores regardless of sex. (F) Ejection fraction (EF) increased with age in males but not females. (G, H) The increase in EF was closely graded by the levels of frailty in males but this relationship did not exist in females. (I) E/A ratios declined with age in males only. (J, K) The age-related decline in E/A ratio was graded by frailty in males but not in females. Age and sex effects were assessed with 2-way repeated-measures ANOVA or a mixed effects model as appropriate, with Sidak’s multiple comparisons post hoc tests as described in the Method. Correlations were evaluated with a Pearson’s r test. The * denotes significant effect of age (p < .05). Values of n = 13 mice per group for HR and EF; values for the E/A ratios were 11 middle-aged (MA) males, 11 MA females, 8 older males, and 8 older females.

Figure 2.

Age-related cardiac remodeling differs between the sexes and is graded by frailty in a sex-specific fashion. (A) Left ventricular internal diameter in diastole (LVIDd) was unchanged by age but was higher in older females than older males. (B, C) LVIDd was not related to frailty scores in either males or females. (D–F) There was no effect of age, sex, or frailty on left ventricular posterior wall in diastole (LVPWd). (G) There was a significant effect of age on intraventricular septum in diastole (IVSd). (H, I) The age-associated increase in IVSd thickness was closely graded by frailty index (FI) scores in males and females. (J) There was a significant increase in left ventricular (LV) mass with age. (K, L) The age-associated increase in LV mass was graded by the level of frailty in females but this pattern was not observed in males. Data were analyzed with 2-way repeated-measures ANOVA with age and sex as main factors, using Sidak’s multiple comparisons post hoc tests as described in the Method. Correlations with frailty were accomplished with a Pearson’s r test. The * denotes significant effect of age or sex (p < .05). Values of n = 13 mice per group. MA; middle-aged.

Figure 3.

Ventricular collagen levels increase with age in males only and this increase is graded by the level of frailty in males, but not females. (A) The levels of hydroxyproline, a major component of collagen, increased with age in males only and were significantly higher in females than males in the younger group. (B, C) Collagen content was graded by frailty in hearts from males but not females. Data were analyzed with a 2-way ANOVA (Sidak’s multiple comparisons post hoc test) and correlations were assessed with a Pearson’s r. The * denotes significant effect of age or sex (p < .05). Values of n = 4–8 middle-aged males, 5 middle-aged females, 7–14 older males, and 10 older females.

Figure 4.

Effects of age and sex on serum levels of inflammatory cytokines and chemokines. (A–N) There were few effects of age on serum cytokines and chemokines, although levels of the pro-inflammatory cytokines IL-3, IL-5, and GM-CSF did decline with age in both sexes. There was also a significant effect of sex on IL-12p40 levels, with 23-month-old females having higher levels than age-matched males. (O–T) In males only, levels of the chemokine eotaxin increased with age, and at 23 months were significantly higher in males than females. Inversely, levels of RANTES were higher in females than males at 23 months. Levels of KC decreased with age, but there was no effect of age or sex on MCP-1, MIP-1α, or MIP-1β. (U–W) Levels of the anti-inflammatory cytokine IL-10 declined with age, and there was a significant effect of sex such that females had higher levels than males. There was no effect of age or sex on serum levels of the other anti-inflammatory cytokines (IL-4 or IL-13). Age and sex effects were assessed with mixed effects models with Sidak’s multiple comparisons post hoc tests as described in the Method. Values of n = 5 middle-aged (MA) males, n = 13 old males, n = 6 MA females, and n = 7 old females.

Figure 5.

Multiple age-associated changes in myocardial structure and function are closely graded by serum cytokine levels in male mice. (A, B) Ejection fraction (EF) was inversely proportional to levels of the pro-inflammatory cytokines IL-3 and GM-CSF. (C) Left ventricular (LV) mass increased in proportion to levels of IL-6. (D) E/A ratios increased as serum levels of IL-3 increased. (E–G) An increase in left ventricular posterior wall thickness in systole (LVPWs) was associated with higher levels of IL-1β, eotaxin, and TNF-α. (H–J) Intraventricular septum in diastole (IVSd) was positively associated with Il-1α, IL-1β, and MIP-1α. The correlations were conducted with a Pearson’s r (p < .05). Values of n = 5 middle-aged and n = 13 old male mice for all measures except E/A ratios where n = 5 middle-aged and n = 8 old.