Short-term high-fat diet impairs anterograde and retrograde memory consolidation, but not retrieval in aged rats

Abstract

Background:

Aging increases vulnerability to cognitive decline, and ultraprocessed diets high in saturated fat may accelerate this trajectory. Although short-term high-fat diet (HFD) exposure is known to impair memory in aged animals, the specific stages of memory most susceptible to short-term HFD remain unclear.

Methods:

This study examined how short-term HFD influences anterograde consolidation, retrograde consolidation, and retrieval of long-term fear memory in aged rats. Male F344 × BN F1 rats (22–24 months) consumed chow or three days of HFD provided at distinct times relative to contextual and cued fear conditioning to isolate each memory phase. Importantly, this brief HFD protocol minimizes metabolic disturbances typically produced by longer-term diet manipulation, allowing us to isolate the effects of macronutrient composition on memory processes.

Results:

Three days of HFD before or immediately after conditioning significantly impaired contextual and cued fear memory, reflecting disrupted anterograde and retrograde consolidation. In contrast, three days of HFD before retrieval had no effect on memory performance.

Conclusion:

These findings demonstrate that short-term consumption of ultraprocessed HFD selectively impairs consolidation while sparing retrieval of hippocampal- and amygdala-dependent memory in aging. These findings are important because identifying the specific memory processes that are disrupted, rather than global memory dysfunction, helps narrow mechanistic targets and informs the development of more precise interventions to mitigate diet-related cognitive decline in aging.

Background

Diet plays an essential role in health and disease across the lifespan, with balanced, nutrient-rich diets protecting against cardiovascular diseases, type 2 diabetes, gastrointestinal diseases, cancers, and metabolic dysfunction (Collaborators GBDD, 2019; Mozaffarian et al., 2015; World Health Organization, 2020), and unhealthy diets exacerbating these same conditions (Beyaz et al., 2016; Mana et al., 2021; Mihaylova et al., 2023). Growing evidence further demonstrates that diet profoundly influences brain health and cognition, with older individuals exhibiting heightened vulnerability compared to younger adults (Butler et al., 2026; Butler et al., 2021; Butler et al., 2025; Gomez-Pinilla, 2008; Gonzalez Olmo et al., 2021; Madison et al., 2020; Ramirez-Salazar et al., 2021; Spencer et al., 2017). Despite this growing recognition, the specific memory processes disrupted by unhealthy dietary exposure, particularly in aging, remain incompletely understood.

Much of what is known about diet-related cognitive impairments derives from animal models of diet-induced obesity. While these models reliably demonstrate memory deficits, they have offered limited insight into underlying neural mechanisms because obesity is accompanied by widespread systemic pathology including metabolic dysregulation, chronic inflammation, and gut microbiome alterations (Bluher, 2025; Butler et al., 2025; Chen et al., 2006; Guo et al., 2020; Xiong et al., 2022). These comorbidities make it difficult to isolate the direct effects of diet on memory processes themselves.

To circumvent these confounds, we have employed a short-term, three-day high-fat diet (HFD) paradigm that does not produce obesity or overt metabolic dysfunction (Butler et al., 2025). Using this approach, we previously demonstrated that HFD exposure selectively impairs long-term contextual and cued fear memory in aged, but not young adult, rats, while leaving short-term memory and learning intact (Spencer et al., 2017). These findings indicate that HFD disrupts memory processes occurring after encoding. Consistent with this interpretation, prior studies showed that HFD-induced memory impairments were accompanied by synaptic plasticity deficits and rapid increases in neuroinflammation within the hippocampus and amygdala, and that blocking interleukin-1 signaling prevented both synaptic plasticity and behavioral deficits (Gonzalez Olmo et al., 2023; Spencer et al., 2017). At the cellular level, short-term HFD selectively suppressed late-phase long-term potentiation (L-LTP), while sparing early-phase LTP, further supporting a specific disruption of consolidation-related mechanisms (Abraham, 2003; Barco et al., 2005; Frey and Morris, 1997).

Importantly, in all of our prior studies, HFD exposure occurred before learning, meaning that observed deficits were necessarily interpreted as impairments in anterograde memory consolidation. However, memory formation is a multistage process, and whether HFD also disrupts retrograde consolidation or retrieval has not been directly tested. This distinction is critical, as consolidation and retrieval rely on partially distinct neural circuits and molecular mechanisms, and are differentially vulnerable to aging and neuroinflammatory perturbations (Barrientos et al., 2006; Barrientos et al., 2002a; Barrientos et al., 2004; Constantinou et al., 2024; Glisky, 2007; Schafe and LeDoux, 2000; Schafe et al., 1999; Wais et al., 2017).

The present study was designed to address this gap by systematically dissociating the effects of HFD on anterograde consolidation, retrograde consolidation, and retrieval in an aging model. To this end, animals consumed the three-day HFD either prior to fear conditioning, immediately after conditioning, or prior to retrieval testing. We report that HFD selectively impairs both anterograde and retrograde consolidation, while sparing retrieval. These findings refine our understanding of diet-related memory vulnerability in aging by identifying consolidation, rather than retrieval, as the key target for intervention following acute dietary insult.

Materials and methods

Subjects

Aged (22–24 months) male F344 × BN F1 rats were obtained from the National Institute on Aging rodent colony maintained by Charles River. Female rats of this strain were not available from this or any other vendor at the time these studies were completed. Therefore, they are not included here but will be included in future studies as they become available. F344 × BN F1 rats are commonly used in aging research as they remain relatively healthy and exhibit preserved baseline cognitive function at advanced ages compared to young adults (Barrientos et al., 2006), enabling the study of age-related vulnerability independent of senility or severe aging-associated pathology while avoiding floor effects. Upon arrival, rats weighed approximately 550 g. Rats were housed in condition-matched pairs in ventilated cages with corn cob bedding, and had unrestricted access to food and water. The colony room was maintained at 22 ± 1 °C on a 12-hour light/dark cycle (lights on at 07:00 h). All animals were given at least one week to acclimate before beginning experimental procedures. Rats were randomly assigned to experimental groups across all three experiments. Separate cohorts of rats were used for each experiment (exp. 1: n = 8 rats per group, exp. 2: n = 7–8 rats per group, and exp. 3: n = 7–8 rats per group) for a total of 46 rats. All experiments were conducted in accordance with the protocol (2018A00000063-R2) approved by the Ohio State University Animal Care and Use Committee. Every effort was made to minimize the number of animals used and their suffering.

Diet

Animals were assigned to one of two diet conditions: standard grain-based chow (Teklad Diets, TD.8640; energy density 3.0 kcal/g, providing 29% of calories from protein, 54% from carbohydrates [no added sweetener], and 17% from fat [0.9% saturated, 1.2% monounsaturated, 2.7% polyunsaturated]) or a purified HFD (Inotiv TD.06414, energy density 5.1 kcal/g, providing 18.4% of calories from protein, 21.3% from carbohydrates [90 g/kg sucrose, 160 g/kg maltodextrin], and 60.3% from fat [37% saturated, 47% monounsaturated, 16% polyunsaturated]). Body weights were recorded throughout each experiment between 9:00–10:00a.m.

Contextual fear conditioning

Hippocampus- and amygdala-dependent long-term memory was assessed using contextual fear conditioning (CFC). Conditioning was conducted in Coulbourn Instruments Habitest Modular chambers (12″ W × 10″ D × 12″ H) equipped with two solid metal and two Plexiglass walls, a ceiling-mounted light and speaker, and a removable grid floor connected to a shock generator providing eight independent outputs. Each chamber was enclosed in a sound-attenuating cubicle (23″ W × 20″ D × 24″ H) and cleaned with water and 70% ethanol between sessions.

Each rat was removed from its home cage and placed individually into a conditioning chamber, where it was allowed to explore for 2 min before being presented with a 15-second auditory tone (76 dB) immediately followed by a 2-second foot shock (1.5 mA). To detect any signs of illness or lethargy, freezing activity was monitored during the exploration period during conditioning and again during the pre-tone period on test day. Following shock termination, rats were promptly returned to their home cages. 9 days later, contextual and cued fear memory were tested to assess hippocampus- and amygdala-dependent memory, respectively (Kim and Fanselow, 1992; Phillips and LeDoux, 1992; Rudy et al., 2002). During the contextual test, rats were placed back into the original conditioning chamber for 6 min, and freezing behavior was recorded. For the cued-fear test, rats were exposed to a distinct context featuring novel visual, and tactile cues (e.g., wire walls, cob bedding, dim lighting). Freezing was scored for 3 min before the tone and 3 min during tone presentation. Freezing was defined as complete immobility accompanied by shallow breathing and autonomic responses such as piloerection. Behavior was scored manually, by at least 2 researchers, in real time using a 10-second interval time-sampling procedure, starting 10 s after the rat was placed in the chamber. Freezing data are expressed as the percentage of total observation time spent freezing.

Experimental design

This study examined how HFD influences different stages of long-term memory, including anterograde consolidation, retrograde consolidation, and retrieval. Diet consumption was manipulated to occur at specific time points corresponding to each stage. Three experiments were conducted using separate cohorts of rats for each. In Experiment 1, the effects of HFD on anterograde consolidation was explored (Fig. 1A). Aged rats received chow or HFD for three days prior to conditioning. Immediately after conditioning (day 1), all rats were returned to chow. Long-term contextual memory was assessed nine days later (day 10), followed two hours later by cued memory testing. In Experiment 2, the effects of HFD on retrograde consolidation was examined (Fig. 2A). To assess how HFD influences consolidation after learning, rats received chow or HFD immediately after conditioning and continued for three days (days 1–3). Contextual memory was evaluated on day 10, and cued memory was tested two hours later. In Experiment 3, the effects of HFD on retrieval was determined (Fig. 3A). All rats consumed chow before and after conditioning (days 1–7). To isolate effects on retrieval, HFD was introduced three days before contextual and cued memory testing (days 8–10). Contextual memory was measured on day 10, followed by cued memory two hours later.

Fig. 1.

Effects of HFD on Anterograde Consolidation in F344 × BN rats (n = 8 per group). (A) Experimental timeline outlining dietary exposure and behavioral testing. (B) Percent change in body weight across time. (C) Freezing during conditioning. (D) Freezing during contextual memory test. (E) Freezing during cued memory test: (left) in novel chamber prior to tone presentation, and (right) during tone presentation. Data are presented as mean ± SEM. *p < 0.05; **p < 0.01.

Fig. 2.

Effects of HFD on Retrograde Consolidation in F344 × BN rats (n = 7–8 per group). (A) Experimental timeline outlining dietary exposure and behavioral testing. (B) Percent change in body weight across time. (C) Freezing during conditioning. (D) Freezing during contextual memory test. E) Freezing during cued memory test: (left) in novel chamber prior to tone presentation, and (right) during tone presentation. Data are presented as mean ± SEM. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

Fig. 3.

Effects of HFD on Retrieval in F344 × BN rats (n = 7–8 per group). (A) Experimental timeline outlining dietary exposure and behavioral testing. (B) Percent change in body weight across time. (C) Freezing during conditioning. (D) Freezing during contextual memory test. (E) Freezing during cued memory test: (left) in novel chamber prior to tone presentation, and (right) during tone presentation. Data are presented as mean ± SEM. *p < 0.05; ****p < 0.0001.

Statistical analysis

For all experiments, 8 rats per group were used. In experiment 2, one cage of chow-fed rats (n = 2) experienced a technical error in which the tone did not turn on during conditioning. Therefore, their scores were excluded from the analysis of the cued fear test. Statistical analyses were performed using Prism version 10. Grubb’s analyses were run to detect statistical outliers. Depending on the outcome measure, behavioral data were analyzed with one-way ANOVAs or unpaired t-tests with Welch’s correction. When significant effects were detected in the one-way ANOVAs, Šídák’s post hoc multiple comparisons tests were conducted to determine pairwise group differences. Statistical significance was defined as α = 0.05.

Results

Short-term HFD impairs anterograde consolidation in aged rats

Body weight was tracked across the duration of the experiment (day −2 through day 10). At baseline on day −2, when animals were first introduced to the HFD, body weights were comparable between groups (Welch’s t(13.09) = 2.288e-009, p > 0.9999; Fig. 1B). Thereafter, rats fed the HFD gained more than chow-fed controls on several days of measurement, including days −1, 0, 1, 3, 5, and 8 (day −1: Welch’s t(8.907) = 4.698, p < 0.01; day 0: Welch’s t(9.188) = 4.219, p < 0.01; day 1: Welch’s t(7.403) = 5.169, p < 0.01; day 3: Welch’s t(7.503) = 4.344, p < 0.01; day 5: Welch’s t(8.575) = 3.538, p < 0.01; day 8: Welch’s t(7.878) = 2.992, p < 0.05). By day 10, body weight differences were no longer statistically significant (Welch’s t(8.843) = 2.262, p > 0.05).

Freezing during conditioning was not significantly different across groups (Welch’s t(12.71) = 0.2774, p > 0.05; Fig. 1C). Despite similar acquisition, HFD-fed rats froze significantly less during contextual memory testing than chow-fed controls (Welch’s t(11.95) = 4.002, p < 0.01; Fig. 1D). A one-way ANOVA showed significant differences in freezing between aged rats fed chow and HFD during cued memory testing when the tone was present and absent (F(3, 28) = 19.37, p < 0.0001; Fig. 1E). Šídák’s post hoc comparisons confirmed that aged HFD-fed rats froze significantly less than chow-fed rats only while the cue was present (p < 0.01; Fig. 1E).

Short-term HFD impairs retrograde consolidation in aged rats

Across days 1–10, body weights again increased in HFD-fed rats. On day 1, the groups did not differ (Welch’s t(13.35) = 4.415e-009, p > 0.9999; Fig. 2B), but by days 2, 3, 4, 6, 8, and 10, HFD-fed rats showed consistently greater body weight (day 2: Welch’s t(13.07) = 3.666, p < 0.01; day 3: Welch’s t(10.87) = 9.199, p < 0.0001; day 4: Welch’s t(7.190) = 4.673, p < 0.01; day 6: Welch’s t(10.75) = 5.011, p < 0.001; day 8: Welch’s t(8.828) = 4.038, p < 0.01; day 10: Welch’s t(11.49) = 2.252, p < 0.01).

During conditioning, freezing behavior did not differ between groups (Welch’s t(9.518) = 1.000, p > 0.05; Fig. 2C). In contrast, deficits emerged during contextual memory testing. Aged rats that were fed HFD froze significantly less than chow-fed rats (Welch’s t(13) = 2.88, p < 0.05; Fig. 2D). HFD also impacted performance during the cued memory test. A one-way ANOVA indicated significant differences across tone-present and tone-absent periods (F(3, 25) = 50.03, p < 0.0001), and Šídák’s comparison demonstrated a pronounced reduction in freezing among HFD-fed rats only during tone presentation (p < 0.0001; Fig. 2E).

Short-term HFD does not affect retrieval in aged rats

Body weight remained similar across groups early in the experiment (days 1–7), with no significant differences detected (day 1: Welch’s t(8.237) = 2.196, p > 0.05; day 2: Welch’s t(12.84) = 0.8524, p > 0.05; day 3: Welch’s t(12.35) = 0.2338, p > 0.05; day 5: Welch’s t(12.61) = 1.248, p > 0.05; day 7: Welch’s t(9.954) = 0.3669, p > 0.05; Fig. 3B). Once the HFD was introduced, body weight increased significantly in the HFD group (day 8: Welch’s t(10.58) = 2.656, p < 0.05; day 9: Welch’s t(12.39) = 5.759, p < 0.0001; day 10: Welch’s t(12.34) = 7.378, p < 0.0001).

Freezing during conditioning did not differ between groups (Welch’s t(12.71) = 1.387, p > 0.05; Fig. 3C). Importantly, contextual memory was also intact as HFD-fed and chow-fed rats performed similarly (Welch’s t(8.574) = 0.376, p > 0.05; Fig. 3D). Although the one-way ANOVA for cued memory revealed an overall effect across tone conditions (F(3, 24) = 26.28, p < 0.0001), a post hoc comparison showed no diet-related difference in freezing during the tone presentation (p > 0.05; Fig. 3E), indicating that memory retrieval remained preserved in the presence of HFD.

Discussion

The present study sought to identify which phases of long-term memory are most vulnerable to disruption by HFD exposure in aging. Building on our prior work demonstrating that short-term HFD does not impair learning (Spencer et al., 2017), we temporally targeted dietary manipulation to specific memory stages to determine whether previously reported deficits reflect impaired consolidation, disrupted retrieval, or both. Using contextual and cued fear conditioning, we show for the first time that HFD selectively impairs both anterograde and retrograde consolidation, while sparing memory retrieval. These findings identify consolidation as the memory phase most sensitive to acute dietary insult in aged rats.

Importantly, groups did not differ in body weight prior to dietary manipulation, and the modest (~4%) increase observed following HFD is consistent with our prior work using this paradigm (Butler et al., 2026; Butler et al., 2025; Butler et al., 2023; Spencer et al., 2017). A key strength of the short-term HFD protocol is that it minimizes obesity and major metabolic disruption (Butler et al., 2025), allowing isolation of dietary composition as a causal factor. Under these conditions, we show that brief exposure to an ultraprocessed, saturated fat-rich diet is sufficient to impair consolidation of long-term memory, independent of overt metabolic pathology.

The consolidation deficits observed here closely mirrors effects we have reported repeatedly using the same dietary paradigm (Butler et al., 2026; Butler et al., 2025; Butler et al., 2023; Spencer et al., 2017). In prior studies, short-term HFD consumption in aged rats rapidly elevated neuroinflammatory signaling within the hippocampus and amygdala, and blocking interleukin-1 signaling prevented both synaptic and behavioral impairments (Gonzalez Olmo et al., 2023; Spencer et al., 2017). These neuroimmune changes were causally linked to suppression of L-LTP, a cellular correlate of memory consolidation (Abraham, 2003; Bramham and Messaoudi, 2005; Frey and Morris, 1997), while early-phase LTP (E-LTP) and short-term memory remained intact. Given the robustness and reproducibility of this mechanistic relationship, the present study was not designed to replicate neuroinflammatory measurements, but rather to address a distinct and unresolved question. Specifically, whether retrieval processes are also compromised by HFD. Accordingly, physiological measures were limited to body weight as a straightforward index of dietary consumption, allowing focused testing of memory-phase specifically without unnecessary duplication of established mechanistic findings.

Similarly, we did not re-examine brain-derived neurotrophic factor (BDNF) signaling in the present study. Prior work using this same short-term HFD paradigm has demonstrated suppression of BDNF-dependent synaptic plasticity in aged rats, consistent with selective disruption of consolidation-related mechanisms (Gonzalez Olmo et al., 2023). Rather than reassessing previously established molecular correlates, the present experiments were designed to leverage temporal manipulation of diet exposure to functionally dissociate anterograde consolidation, retrograde consolidation, and retrieval. Future studies integrating parallel molecular and neurotrophic analyses with this temporal framework will be important for further refining the biological mechanisms linking acute dietary insults to specific memory processes.

The finding that short-term HFD impaired retrograde consolidation is particularly notable given the temporal dynamics of memory stabilization. Memory consolidation remains labile for about 24 h following learning, during which molecular and synaptic processes are highly sensitive to perturbation (Barrientos et al., 2002b; Bekinschtein et al., 2007; Dudai, 2004; Mondadori et al., 1991). In the present study, HFD exposure initiated immediately after conditioning overlapped only partially with this consolidation window, yet was sufficient to impair long-term memory. These results suggest that even one or two days of unhealthy dietary exposure during critical post-encoding periods may be sufficient to disrupt consolidation in aging.

In contrast, memory retrieval remained intact. Retrieval of consolidated memories increasingly relies on coordinated activity within distributed cortical networks as memories undergo systems consolidation and become less dependent on hippocampal and amygdalar involvement. Lesion and inactivation studies show that while hippocampal and amygdalar circuits are essential for recent memory formation and consolidation, retrieval at later time points can be supported by neocortical regions such as the anterior cingulate cortex, retrosplenial cortex, lateral prefrontal cortex, and ventral parietal cortex (Anagnostaras et al., 1999; Berlau and McGaugh, 2003; Bontempi et al., 1999; Frankland and Bontempi, 2005; Frankland et al., 2004; Keene and Bucci, 2008a, b; Kim and Fanselow, 1992; McGaugh et al., 1996; McGaugh et al., 2002; Sridhar et al., 2023; Vann et al., 2009). Because retrieval testing in the present study occurred nine days after conditioning, memory expression would be expected to rely primarily on these distributed cortical circuits. The redundancy and distributed nature of these networks likely confer resilience against the transient inflammatory and synaptic perturbations induced by HFD, providing a plausible explanation for preserved retrieval despite impaired consolidation.

Together, these findings have important implications for understanding how aging-related vulnerabilities interact with dietary composition to influence cognitive function. Aging is associated with heightened neuroinflammatory tone and reduced synaptic plasticity, particularly within hippocampal and amygdalar circuits that support memory consolidation (Barrientos et al., 2015). Our results demonstrate that even brief exposure to an unhealthy diet can selectively disrupt these consolidation processes in aged organisms, independent of major metabolic dysfunction. By identifying consolidation, rather than retrieval, as the primary interventional target of acute dietary insult, this work highlights dietary quality as a modifiable factor with the potential to influence cognitive resilience later in life.

Conclusion

Short-term HFD selectively impairs anterograde and retrograde consolidation while sparing retrieval. These findings identify consolidation as the memory phase most vulnerable to acute nutritional disruption and underscore the sensitivity of hippocampal- and amygdalar-dependent processes to saturated-fat overconsumption. Collectively, the results suggest that even brief unhealthy dietary exposure may heighten vulnerability to long-term memory impairments in aging by disrupting consolidation processes that are central to cognitive health across the lifespan.