Introduction

The incidence of age-related metabolic disease, driven by fat accumulation in the liver, nearly doubles from 45 to 65 years of age (Harris et al. 1998; Cowie et al. 2009; CDC 2014, 2017). The prevalence of nonalcoholic fatty liver disease (NAFLD) rises from young adulthood to middle age, with a prevalence exceeding 40% in people over 70 years of age (Frith et al. 2009; Wang et al. 2013; Bertolotti et al. 2014). Excess hepatic lipid content is associated with impaired glucose tolerance (Borel et al. 2015) and an increased risk of develo** Type II diabetes mellitus (T2DM) (Li et al. Full size image

As a first step to better understand the CR and age effects on hepatic lipid content, we next assessed liver ATP citrate lyase (ACLY) and acetyl CoA carboxylase (ACC) mRNA expression. These are two key enzymes involved in hepatic de novo lipogenesis (Thampy and Wakil 1988; Kim et al. 2017). CR did not affect mRNA expression of ACC or ACLY at 10-, 18-, or 26- months of age (Fig. 3C, D). However, CR decreased ACC mRNA expression at 28 months of age (Fig. 3D, P  < 0.01).

Lifelong moderate CR improves glucose clearance at 10 and 18 months of age

Because hepatic lipid accumulation is tightly coupled to glucose homeostasis (Borel et al. 2015; Lomonaco et al. 2016), we next performed oral glucose tolerance tests (OGTT) to assess the effects of calorie restriction on glucose clearance and oral glucose-stimulated insulin. Moderate calorie restriction improved oral glucose clearance in 10- and 18-month-old mice, the same mice in whom calorie restriction had lowered liver lipid concentration (Fig. 4A, C and I, P < 0.0001). Interestingly, coinciding with decreased liver triglycerides, glucose clearance was improved in ad libitum fed 26- and 28-month-old mice relative to 10- and 18-month-old mice. CR did not further improve glucose clearance in 26- and 28-month-old mice compared to ad libitum fed age-matched controls (Fig. 4E, G and I).

Fig. 4
figure 4

Oral glucose clearance and oral glucose-stimulated insulin secretion (OGSIS). Blood glucose curves (A, C, E, and G) and the area-under-the-curve (I) from oral glucose tolerance tests in ad libitum fed and 15% Calorie Restricted (CR) mice at 10 (n = 15), 18 (n = 24–25), 26 (n = 13–26), and 28 months of age (n = 9–10). a,b,cLetters that differ indicate differences, P < 0.05; two-way ANOVA with Tukey’s adjustment for multiple comparisons. Oral glucose-stimulated insulin secretion (OGSIS) (B, D, F, H) from baseline to 15 min after oral glucose (2.5 g/kg BW) gavage (10 months: n = 15; 18 months: n = 21–24; 26 months: n = 12–25; 28 months: n = 6–9). P values reflect results of paired samples t tests to assess the change in serum insulin concentration between timepoints within each mouse (B, D, F, H). Effect of aging on OGSIS 15-min-post oral glucose gavage (J); a,b,cLetters that differ indicate differences, P < 0.05; two-way ANOVA with Tukey’s adjustment for multiple comparisons. Data presented as Mean ± SEM

Full size image

Having demonstrated that CR improves glucose clearance in 10- and 18-month-old mice and that glucose clearance improves in advanced age, we sought out to determine if these improvements were a result of changes in oral glucose stimulated insulin secretion (OGSIS). We assessed serum insulin concentrations at baseline and 15 min after oral glucose was administered. In each age and diet group, oral glucose gavage increased serum insulin concentrations 15 min after administration (Fig. 4B, D, F, H). CR decreased OGSIS only at 18 and 28 months of age. Of note, aging increases oral glucose stimulated serum insulin in both ad libitum and CR mice (P < 0.001, Fig. 4J).

Lifelong moderate CR increases insulin sensitivity in 28-month-old mice

Having established the effect of aging and CR on glucose clearance and OGSIS, we sought out to assess the effect of CR on insulin sensitivity in mice of advance age (28 months of age). Although basal glucose was nearly identical in both groups of mice, insulin more severely decreased blood glucose concentration in calorie restricted mice (Fig. 5A and B). We next collected tail blood at 0 and 15 min during the ITT to assess the effect of calorie restriction on hypoglycemia-stimulated glucagon secretion. Glucagon, the counterregulatory hormone to insulin, increases in response to hypoglycemia and extended fasting to promote hepatic glycogenolysis and increase blood glucose (Stern et al. 2019; Vasileva et al. 2022). While serum glucagon did not change from 0 to 15 min after insulin injection in ad libitum mice, we observed a robust rise in serum glucagon in the calorie restricted group (P = 0.004, Fig. 5C). This more robust rise in serum glucagon would be expected to limit the apparent insulin sensitivity measured by ITT. In turn, the CR induced improvement in insulin sensitivity is likely more robust than suggested by the ITT.

Fig. 5
figure 5

Lifelong 15% calorie restriction improves insulin sensitivity in advanced age. Blood glucose curves (A) and area-under-the-curve (B) from insulin tolerance tests in 28-month-old ad libitum fed and 15% calorie restricted (CR) mice (n = 7–8); P values reflect results of unpaired t-tests to assess difference between diet groups. C Serum glucagon concentrations at baseline and 15 min after intraperitoneal insulin injection (0.25 IU/kg BW) (n = 6–8); P values reflect results of paired samples t tests to assess the change in serum insulin concentration between timepoints within each mouse. Data presented as Mean ± SEM

Full size image

Lifelong moderate CR improves physical function

Because metabolic health correlates with physical function (Fritschi et al. 2017), we next assessed the effects of caloric restriction and aging on measures of physical function. We performed grip strength tests and found that calorie restriction improved all limb grip strength at 10- and 18-months of age and forelimb grip strength at 18-months of age (Fig. 6A  and B, P < 0.001 and P < 0.0001, respectively), but we saw no effect of CR on grip strength at 26 or 28 months of age. To assess balance and coordination, we performed the rotarod test in 26- and 28-month-old mice. Calorie restriction tended to increase the time to fall in 26-month-old mice (P = 0.07, Fig. 6C) and significantly increased time to fall in 28-month-old mice (Fig. 6C, P = 0.004).

Fig. 6
figure 6

In vivo measures of physical function. All limb (A) and forelimb (B) grip strength in ad libitum fed and 15% calorie restricted mice at 10 (n = 15), 18 (n = 23–25), 26 (n = 13–24), and 28 (n = 5–9) months of age; a,b,cLetters that differ indicate differences, P < 0.05; two-way ANOVA with Tukey’s adjustment for multiple comparisons (A, B). Rotarod task was performed in 26- (n = 6–10) and 28-month-old mice (n = 9); P values reflect results of unpaired t-tests to assess difference between diet groups. Data presented as Mean ± SEM

Full size image