Behavioral effectsLong-term and short-term HFD induce anxiety-like behavior in aged male rats

In the OFT, there were main effects of age (F(1,24) = 20.34, p = 0.0001), diet (F(1,24) = 16.52, p = 0.0004), and an age x diet interaction (F(1,24) = 23.15, p < 0.0001; Fig. 2A, D) for time spent in the center. A pairwise analysis showed that aged chow-fed rats spent more time in the center of the OFT compared to young adult chow-fed controls (p < 0.0001) (Fig. 2A). However, long-term HFD caused a significant reduction in time spent in the center in aged rats compared to chow-fed aged controls (p < 0.0001) (Fig. 2A). A 2-way ANOVA of freezing behavior revealed main effects of age (F(1,25) = 11.95, p = 0.0020) and diet (F(1,25) = 8.711, p = 0.0068), and an interaction effect (F(1,25) = 6.489, p = 0.0174; Fig. 2B). Pairwise comparisons showed that aged HFD-fed rats spent significantly more time freezing compared to HFD-fed young adult and chow-fed aged controls (p = 0.0028 and p = 0.0024, respectively (Fig. 2B). For rearing, a main effect of diet (F(1,25) = 9.870, p = 0.0043) and an age x diet interaction effect (F(1,25) = 7.188, p = 0.0128; Fig. 2C) were observed. Aged, HFD-fed rats had significantly fewer rearing episodes compared to HFD-fed young adult and chow-fed aged controls, evidenced by pairwise analysis (p = 0.0016 and p = 0.0284, respectively). In all three of these measures, young adult rats, regardless of diet condition, did not differ from each other (p > 0.05).

Fig. 2

Measures of anxiety-like behavior via the open field test (OFT) in young and aged rats following long- and short-term HFD consumption. Percentage of time spent in center during OFT following (A) long and (E) short-term HFD consumption. Percentage of time spent freezing during OFT following (B) short and (F) long-term HFD consumption. The number of rearing episodes during the OFT following (C) long and (G) short-term HFD consumption. Representative plots of routes traced during OFT following (D) long and (H) short-term HFD consumption. *p < 0.05; **p < 0.01; *** p < 0.001; ****p < 0.0001

These anxiety-like behavioral changes emerge rapidly after HFD onset, as similar effects were observed following short-term HFD. At this timepoint, 2-way ANOVA assessment showed main effects of age (F(1,19) = 29.10, p < 0.0001) and diet (F(1,19) = 22.21, p = 0.0001), as well as an interaction effect (F(1,19) = 22.70, p < 0.0001) (Fig. 2E, H). Consistent with the long-term cohort, aged chow-fed rats spent significantly more time in the central part of the OFT chamber compared to chow-fed young controls (p < 0.0001). This aging-related anxiolytic effect was prevented by HFD, as aged HFD-fed rats spent significantly less time in the center than their chow-fed controls (p < 0.0001) (Fig. 2E). For freezing behavior, a main effect of age (F (1,19) = 19.67, p = 0.0003), diet (F(1,19) = 26.29, p < 0.0001), and an age x diet interaction effect (F(1,19) = 13.03, p = 0.0019) were observed (Fig. 2F). Pairwise analyses revealed that aged HFD-fed rats froze significantly more than their young adult and chow-fed counterparts (p < 0.0001 and p = 0.0001, respectively). In the rearing measure, a 2-way ANOVA showed main effects of age (F(1,19) = 4.582, p = 0.0455) and diet (F(1,19) = 7.989, p = 0.0108), and an age x diet interaction effect (F(1,19) = 5.102, p = 0.0358; Fig. 2G). As with long-term HFD, short-term HFD resulted in aged rats having significantly fewer rearing episodes compared to their young adult and chow-fed counterparts (p = 0.0078 and p = 0.0314, respectively). Again, in all three of these measures, young adult rats, regardless of diet condition, did not differ from each other (p > 0.05).

Long-term and short-term HFD induce memory impairments in aged male rats

Following long-term HFD consumption, aged rats experienced significant contextual and cued-fear memory deficits. For contextual memory, a 2-way ANOVA revealed main effects of age (F(1,25) = 5.961, p = 0.0220) and diet (F(1,25) = 24.82, p < 0.0001), as well as an age x diet interaction (F(1,25) = 10.46, p = 0.0034), to decrease time spent freezing, indicating impaired memory (Fig. 3A). Pairwise comparisons revealed that HFD-fed aged rats performed significantly worse on the contextual memory task than HFD-fed young adult and chow-fed aged controls (p < 0.0001 and p = 0.0041, respectively; Fig. 3A). Similar effects were observed in the cued-fear memory task. A 2-way ANOVA showed that 3 months of HFD induced a significant effect of age (F(1,24) = 17.26, p = 0.0004) and diet (F(1,24) = 16.16, p = 0.0005) to reduce time spent freezing, as well as an interaction effect (F(1,24) = 5.029, p = 0.0344; Fig. 3B). Aged rats fed HFD exhibited significant cued-fear memory impairment compared to HFD-fed young adult and chow-fed aged controls (p = 0.0014 and p = 0.0011, respectively; Fig. 3B).

Fig. 3

Long-term memory function assessed via contextual and cued-fear conditioning following long-term and short-term HFD consumption. Freezing behavior during the testing phase for hippocampal-dependent contextual memory following (A) long-term and (C) short-term HFD consumption. Freezing behavior during the testing phase for amygdala-dependent cued-fear memory following (B) long-term and (D) short-term HFD consumption. *p < 0.05; **p < 0.01; ****p < 0.0001

To assess how rapidly these memory impairments occur following initiation of a HFD, we assessed contextual and cued-fear memory function following short-term (just 3 days) consumption of HFD. In the contextual memory task, short-term HFD resulted in main effects of age (F(1,26) = 35.33, p < 0.0001) and diet (F(1,26) = 76.21, p < 0.0001), as well as an age x diet interaction (F(1,26) = 5.731, p = 0.0242), to impair performance on the test (Fig. 3C). Pairwise analysis showed that aged rats fed short-term HFD were significantly impaired to young adult and chow-fed controls (p < 0.0001 and p < 0.0001, respectively; Fig. 3C). On the cued-fear task, short-term HFD induced a main effect of age (F(1,26) = 4.284, p = 0.0485) and an age x diet interaction (F(1,26) = 5.139, p = 0.0319) to reduce time spent freezing (Fig. 3D). Aligning with our previous findings, aged rats fed short-term HFD were significantly impaired compared to young adult and chow-fed controls (p = 0.0284 and p = 0.0303, respectively; Fig. 3D). This finding that just 3 days of HFD induces long-term contextual and cued-fear memory deficits replicates the behavioral findings associated with long-term HFD consumption, indicating that HFD rapidly induces memory impairment which persists throughout diet consumption.

NeuroinflammationLong-term HFD alters inflammatory cytokines in the hippocampus and amygdala of young and aged rats

Following 3 months of either chow or HFD, we measured changes in protein levels of pro and anti-inflammatory cytokines in the hippocampus and amygdala. In the hippocampus, there was a significant age x diet interaction in all cytokines except IL-6 (IL-1β: F(1,25) = 4.441, p = 0.0453; TNF: F(1,26) = 4.544, p = 0.0425; IL4: F(1,25) = 6.132, p = 0.0204; IL10: F(1,24) = 7.211, p = 0.0129; IFNγ F(1,26) = 4.443, p = 0.0448; Fig. 4A-F). Post hoc analyses of these cytokines revealed a significant decrease in cytokine levels in HFD-fed aged rats compared to HFD-fed young adult rats (IL-1β, IL4, and IL10: p < 0.01; TNF, IFNγ p < 0.05). Results of IL-6 showed a main effect of age (F(1,26) = 10.26, p = 0.0036; Fig. 4B), such that aged rats had decreased IL-6, relative to young, regardless of diet condition.

Fig. 4

Inflammatory protein concentration in the hippocampus following long-term HFD consumption in young and aged rats. Levels of (A) IL-1β, (B) IL-6, (C) TNF, (D) IL-4, (E) IL-10, and (F) IFNγ in the hippocampus. *p < 0.05, **p < 0.01

In the amygdala, there was also a significant age x diet interaction for all cytokines except IL-6 and IL-10 (IL-1β: F(1,24) = 4.270, p = 0.0497; TNF: F(1,21) = 11.65, p = 0.0029; IL-4: F(1,24) = 5.169, p = 0.0322; IFNγ: F(1,21) = 12.37, p = 0.0018; Fig. 5). Pairwise comparisons indicated that aged HFD-fed rats had higher levels of IL-1β (p < 0.05; Fig. 5A), TNF (p < 0.005; Fig. 5C) and IL-4 (p < 0.01; Fig. 5D) compared to young HFD-fed rats. Moreover, aged HFD-fed rats had elevated IFNγ levels relative to all other groups (p < 0.005 – 0.05; Fig. 5F). There was no significant differences across groups for IL-6 and IL-10 in the amygdala.

Fig. 5

Inflammatory protein concentration in the amygdala following long-term HFD consumption in young and aged rats. Levels of (A) IL-1β, (B) IL-6, (C) TNF, (D) IL-4, (E) IL-10, and (F) IFNγ in the amygdala. *p < 0.05, **p < 0.01, ***p < 0.001

Short-term HFD alters inflammatory cytokines in the hippocampus and amygdala of aged rats

The concentration of the same proteins were measured in the hippocampus and amygdala of young and aged rats following short-term consumption of either chow or HFD. In the hippocampus, there was a significant age x diet interaction for all cytokines except TNF and IFNγ (IL-1β: F(1,28) = 7.957, p = 0.0087; IL-6: F(1,27) = 10.68, p = 0.0029; IL-4: F(1,28) = 4.656, p = 0.0397; IL-10: F(1,28) = 4.424, p = 0.0446; Fig. 6). Posthoc analyses of each of these cytokines demonstrated a significant increase in cytokines in aged rats fed a HFD, relative to all other groups (p < 0.005 – 0.05). While there was not a significant interaction, there was a main effect of age for TNF levels (F(1,28) = 11.97, p = 0.0017; Fig. 6C), as aged rats displayed higher levels than young. There was also a main effect of diet on IFNγ concentration in the hippocampus (F(1,27) = 12.51, p = 0.0015; Fig. 6F), with HFD-fed rats showing increased levels relative to chow-fed controls.

Fig. 6

Inflammatory protein concentration in the hippocampus following short-term HFD consumption in young and aged rats. Levels of (A) IL-1β, (B) IL-6, (C) TNF, (D) IL-4, (E) IL-10, and (F) IFNγ in the hippocampus. *p < 0.05, **p < 0.01, ***p < 0.001

Additionally, we observed dysregulated cytokines in the amygdala following short-term HFD. There was a significant age x diet interaction for IL-1β (F(1,28) = 6.003, p = 0.0208; Fig. 7A), IL-6 (F(1,19) = 11.04, p = 0.0036; Fig. 7B), and IFNγ (F(1,25) = 9.787, p = 0.0044; Fig. 7F). Tukey’s posthoc tests revealed young HFD-rats had less IL-1β than aged chow- and HFD-fed rats. Furthermore, young HFD-fed rats had increased IL-6 (p < 0.01) and IFNγ (p < 0.05) levels compared to young chow-fed rats, with IFNγ concentration also being increased in young HFD-fed rats relative to aged HFD-fed rats (p < 0.05). For TNF in the amygdala, there was a main effect of age (F(1,28) = 4.326, p = 0.0468; Fig. 7C), as aged rats had increased TNF levels relative to young. Lastly, we observed a main effect of diet on amygdala IL-10 concentration (F(1,28) = 5.008, p = 0.0334; Fig. 7E), with IL-10 levels being increased in HFD-fed rats.

Fig. 7

Inflammatory protein concentration in the amygdala following short-term HFD consumption in young and aged rats. Relative expression of (A) IL-1β, (B) IL-6 (C) TNF (D) IL-4 (E) IL-10, (F) IFNγ levels in the amygdala. ****p < 0.0001

Metabolic effectsLong-term, but not short-term, HFD consumption induces hyperglycemia and visceral adipose tissue inflammation in young and aged rats

Because obesity and metabolic syndrome are associated with significant weight gain, hyperglycemia, and increased inflammation in adipose tissue, we evaluated these metrics at two distinct time points during diet consumption. For the long-term HFD cohort, there was a main effect of diet (F(1,25) = 65.42, p < 0.0001; Fig. 8A) on the percent of body weight gained during the study, such that HFD-fed rats gained more weight than chow-fed rats (p < 0.0001). There was also a main effect of age (F(1,25) = 27.67, p < 0.0001; Fig. 8A) on the percent of body weight gain, where aged rats gained less during the course of the 3mo study, relative to young rats (p < 0.0001). For short-term HFD consumption, there was an age x diet interaction (F(1,19) = 4.936, p = 0.0386; Fig. 8E) for the percent of body weight gained. Post hoc analysis revealed that aged, HFD-fed rats gained more weight than young, HFD-fed rats and young and aged chow-fed controls during the 3 days on the diet (p < 0.05). There were no differences in percent weight gain between young and aged chow-fed rats (p > 0.05).

Fig. 8

Measurement of body weight, circulating metabolic markers, and visceral adipose tissue inflammation in young and aged rats following long- and short-term HFD. A Percent of body weight gained by young and aged rats following 3mo of HFD consumption. Serum levels of (B) fasting glucose and (C) fasting insulin and D) protein levels of IL-1beta in visceral adipose tissue in young and aged rats following 3mo of HFD consumption. E Percent of body weight gained by young and aged rats following 3d of HFD consumption. Serum levels of (F) fasting glucose and (G) fasting insulin and (H) protein levels of IL-1beta in visceral adipose tissue in young and aged rats following 3d of HFD consumption. **p < 0.01; ****p < 0.0001

Following long-term HFD consumption, there was a main effect of diet on serum levels of fasting glucose (F(1,27) = 15.680, p = 0.0005; Fig. 8B) and insulin (F(1,27) = 8.436, p = 0.0073; Fig. 8C) in both young and aged rats. Importantly, there were no differences in fasting glucose or insulin levels in young or aged rats following short-term HFD consumption (Fig. 8F&G). Similarly, only long-term HFD consumption resulted in increased inflammation in visceral adipose tissue, as indicated by increased IL-1β in both young and aged rats that consumed HFD, relative to chow-fed controls (F(1,22) = 54.300, p < 0.0001; Fig. 8D). There were no differences in visceral adipose tissue IL-1β in young or aged rats following short-term HFD consumption (Fig. 8H).

Gut inflammationLong-term, but not short-term, HFD induces a proinflammatory phenotype in the aged colon

Given the intricate gut-brain connection and prior work demonstrating HFD and obesity are each associated with gut inflammation [42,43,44], we measured various pro and anti-inflammatory cytokines in the colon and ileum tissues following both long-term and short-term HFD consumption in young and aged rats. At the 3-month timepoint (long-term) in the colon, a two-way ANOVA indicated a main effect of diet (F(1,24) = 11.27, p = 0.0026; Fig. 9A) on IL-1β such that HFD significantly increased levels compared to chow-fed rats, regardless of age. There was also a significant age x diet interaction for IL-6 (F(1,24) = 5.490, p = 0.0277; Fig. 9B), IL-4 (F(1,25) = 6.339, p = 0.0186; Fig. 9D), and IL-10 (F(1,24) = 5.053, p = 0.0340; Fig. 9E). For all three cytokines, post hoc analysis revealed that young HFD-fed rats had increased levels of each cytokine, relative to all other groups (p < 0.05). There were no differences in colon TNF levels across any condition (p > 0.05). In the ileum, there was a main effect of age for IL-1β (F(1,24) = 5.053, p = 0.0161; Fig. 10A). Specifically, aged rats had lower levels of IL-1β than young rats (p < 0.05). There were no impacts of HFD on the quantified cytokines in the ileum at the 3-month timepoint (Fig. 10). Importantly, there was no impact of short-term HFD on either colon or ileum cytokines that were measured (p > 0.05 for all effects; Figs. 11 and 12, respectively).

Fig. 9

Pro and anti-inflammatory markers in the distal colon of young and aged rats following long-term HFD consumption. Protein levels of (A) IL-1β, (B) IL-6, (C) TNF, (D) IL-4, and (E) IL-10 expressed as pg/g of total protein. *p < 0.05; **p < 0.01; ****p < 0.0001

Fig. 10

Pro and anti-inflammatory markers in the ileum of young and aged rats following long-term HFD consumption. Protein levels of (A) IL-1β, (B) TNF, (C) IL-4, and (D) IL-10 expressed as pg/g of total protein. *p < 0.05

Fig. 11

Pro and anti-inflammatory markers in the distal colon of young and aged rats following short-term HFD consumption. Protein levels of (A) IL-1β, (B) IL-6, (C) TNF, (D) IL-4, and (E) IL-10 expressed as pg/g of total protein

Fig. 12

Pro and anti-inflammatory markers in the ileum of young and aged rats following short-term HFD consumption. Protein levels of (A) IL-1β, (B) IL-6, (C) TNF, (D) IL-4, and (E) IL-10 expressed as pg/g of total protein

MicrobiomeAged rat microbiome is more sensitive to a long-term HFD

We next investigated how long-term HFD impacted the gut microbiota of both young and aged rats. Long-term HFD reduced alpha diversity across both ages, yet older rats exhibited a more pronounced decrease in microbial richness as compared to their young counterparts (Faith’s PD—Fig. 13A and Shannon index—Fig. S1A p < 0.05). HFD also induced robust compositional shifts in the gut microbiome (Fig. S1B). Some of these HFD-induced compositional shifts were independent of age, including HFD-induced expansion of Lactococcus sp, and Blautia sp that occurred in both age groups. Moreover Turicibacter sp. and Clostridia_UCG-014 were downregulated by HFD independent of age (Fig. 13B). However, aged mice also exhibited unique microbiome responses to HFD (Fig. S1C). Notably, long-term HFD led to more pronounced expansion of Ruminococcus torques, Enterococcus sp, Clostridium innocuum, and Eubacterium siraeum in aged rats vs. young rats (Fig. 13C, Age x HFD Two-factor negative binomial (NEGBIN) p < 0.05).

Fig. 13

Three months ingestion of high fat diet (HFD) robustly alters colon microbiome composition dependent of age. A Alpha-diversity indices as measured by Faith’s PD. B Four top bacterial taxa (represented as % total bacteria) upregulated (Lactococcus, Blautia) and downregulated (Turicibacter, Clostridia UCG0-14) by HFD independent of age. C Two factor negative binomial (NEGBIN) reveals age x diet interactions for four bacterial taxa (adjusted p < 0.05). D Heatmap representing Spearman correlation coefficients comparing bacterial genera abundance vs. brain region specific protein levels in the aged group. * p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001

Short-term HFD rapidly alters gut microbiome irrespective of age

As compared to long-term HFD, short-term (3 days) HFD did not change microbial richness as measured by Faith’s PD. Aged rats exhibited lower alpha-diversity compared to their younger counterparts, but only in the chow group (Fig. 14A). Short-term HFD elicited changes to microbial taxa (Fig. S2C), yet most of these effects occurred independent of age. For example, HFD increased Lactococcus and Colidexbacter, and reduced Lachnospiraceae UCG-006 and Prevotellaceae NK3B31 (Fig. 14B and Fig. S2B).

Fig. 14

Three days ingestion of high fat diet (HFD) moderately shifts colon microbiome composition independent of age. A Alpha-diversity indices as measured by Faith’s PD. B Four top bacterial taxa (represented as % total bacteria) upregulated (Lactococcus, Colidextribacter) and downregulated (Lachnospiraceae UCG-006, Prevotellaceae NK3B31) by HFD independent of age. C Two factor negative binomial (NEGBIN) does not indicate a significant age x diet interactions for bacterial taxa. D Heatmap representing Spearman correlation coefficients comparing bacterial genera abundance vs. brain region protein levels in the aged group. * p < 0.05, **p < 0.01, ****p < 0.0001

Spearman correlational analysis in the aged group revealed associations between microbial taxa and neuroinflammation indices after only 3 days on HFD. Most prominent was Lachnospiraceae UCG-006, which was downregulated by age and HFD, and negatively associated with amygdala and hippocampus TNF levels (Fig. 14D). In addition, Prevotellaceae NK3B31, downregulated by HFD, negatively correlated with several inflammatory markers in both the hippocampus and amygdala (Fig. 14D). On the other hand, Lactococcus, upregulated by HFD, positively correlated with all brain inflammatory markers (Fig. 14D).