Abstracts

Rationale: Ketogenic diets (KD) have anticonvulsant effects, but their effects on cognition are less clear. Therefore, we examined the behavioral performance of rats fed a KD on tests of locomotor activity and Pavlovian fear conditioning. We also examined in vivo medial perforant path long-term potentiation (LTP) in rats fed a KD. Methods: Rats were fed a standard diet (SD), a KD, or a calorie-restricted diet (CR) beginning on postnatal day (PD) 21. Rats on the ketogenic diet received 92% of their calories from fat. Rats on the CR served as a control for the slower weight gain experienced by KD fed rats. CR fed rats were given the same chow as the SD but were calorie restricted to allow their weight to match that of the rats fed the KD. On PD 35-60, a 1-hour locomotor activity test and a conditioned fear (contextual or auditory cue) test were performed. On PD 42-45, in vivo median perforant path LTP was assessed by recording field excitatory postsynaptic potentials (fEPSPs) from the dentate gyrus before and after delivering a tetanus. Results: Rats fed the CR and KD showed slower weight gain, but only the KD fed rats had elevated serum β-hydroxybutyrate levels as we have reported previously. KD and SD fed rats did not differ in performance on several activity-related variables (total ambulations, rearing, time at rest). However, CR fed rats exhibited significant hyperactivity relative to the other two groups in terms of total ambulations (whole body movements), particularly during the last 4, 10-minute blocks of testing (p<0.0008). The CR fed rats also reared significantly more often than either of the other two groups during the fourth and fifth time blocks (p<0.0008), and spent significantly less time at rest than the KD fed group (but not SD fed) over the 60 minute test (p=0.002). KD and SD fed groups also performed similarly during the conditioned fear measures of baseline testing, tone-shock training, or during contextual fear or auditory cue testing. Although no significant overall Treatment effects were found in amounts of freezing during the contextual fear test, only the CR fed rats did not show evidence of contextual cue conditioning relative to the pre-test baseline. An ANOVA of the auditory cue data that yielded a significant Treatment by Time interaction, [F(14,280)=2.13, p=0.032], and subsequent significant contrasts showed that, on average across time, the CR fed group froze significantly less often than the KD or SD fed groups (p=0.030 and 0.040, respectively). In vivo LTP in KD fed rats did not differ from rats fed the SD. CR fed rats exhibited LTP, but the magnitude of LTP in CR fed rats was smaller. A two-way repeated measures ANOVA revealed a significant main effect of diet (p = 0.049). Pairwise comparisons revealed that the magnitude of LTP for the CR (p = 0.02) but not the KD (p = 0.30) fed rats was smaller than the SD fed rats. Conclusions: These findings suggest that a KD does not adversely affect the cognitive functions and type of LTP examined. In contrast, a CR may have an adverse effect on the same parameters. (Supported by NIH & JDRF)