Effects of Ketogenic Diet on Behavioral, Metabolic, and Peripheral Tissue Changes in a Propylthiouracil-Induced Hypothyroidism Model

Abstract

Objective

This study aimed to evaluate whether a ketogenic diet (KD) mitigates behavioral, metabolic, and peripheral tissue alterations in a rat model of propylthiouracil (PTU)-induced hypothyroidism.

Subjects and Methods

Adult female Sprague-Dawley rats (n = 28) were randomly assigned to control, KD, PTU, and KD+PTU groups. Hypothyroidism was induced using 0.05% PTU for 6 weeks, while KD was administered during the final 4 weeks. Behavioral testing was conducted during the final phase of the experimental period, followed by biochemical and histological analyses at the end of the 6-week protocol. Behavioral assessments included the open field, light/dark box, forced swim, and Morris water maze tests. Biochemical analyses evaluated serum free thyroxine (T4), thyroid-stimulating hormone (TSH), glucose, β-hydroxybutyrate, tumor necrosis factor alpha (TNF-α), and interleukin 10 (IL-10). Histological evaluations targeted the thyroid gland, liver, white adipose tissue (WAT), and brown adipose tissue (BAT).

Results

PTU significantly reduced serum free T4 levels and impaired spatial learning and mood-related behaviors. KD effectively induced nutritional ketosis and improved systemic glucose metabolism but did not reverse cognitive impairment or depressive-like behaviors associated with hypothyroidism. Histologically, KD attenuated PTU-induced hepatic inflammation, WAT dysfunction, and BAT whitening. Thyroid morphology and circulating TNF-α and IL-10 levels remained unchanged across groups.

Conclusion

KD exerts favorable metabolic and peripheral tissue effects in PTU-induced hypothyroidism but does not rescue associated neurobehavioral deficits, likely due to persistent thyroid hormone deficiency.

Highlights of the Study

• Propylthiouracil-induced hypothyroidism impaired mood-related behavior and spatial learning in female rats.

• Ketogenic diet (KD) induced nutritional ketosis and improved systemic glucose metabolism under hypothyroid conditions.

• KD did not reverse hypothyroidism-related cognitive impairment or depressive-like behavioral alterations.

• Despite persistent suppression of circulating thyroid hormones, the KD attenuated hypothyroidism-induced histopathological alterations in liver and adipose tissues.

Introduction

Hypothyroidism is characterized by insufficient thyroid hormone production and is associated with systemic metabolic disturbances as well as cognitive and mood impairments [1, 2]. Experimental studies have demonstrated that thyroid hormone deficiency disrupts synaptic plasticity, reduces neurogenesis, and alters neurotransmitter homeostasis, leading to measurable neurobehavioral deficits [3, 4]. In addition to central effects, hypothyroidism induces peripheral alterations, including hepatic lipid dysregulation and adipose tissue remodeling [5, 6]. Notably, neurobehavioral symptoms may persist despite restoration of peripheral euthyroidism, highlighting the need for adjunct therapeutic strategies [7].

The ketogenic diet (KD), a high-fat and low-carbohydrate dietary regimen, induces nutritional ketosis by promoting ketone body production. Ketone bodies serve as alternative energy substrates for the brain and have been associated with neuroprotective, anti-inflammatory, and metabolic effects in various disease models [8–10]. KD has been shown to reduce seizure frequency and attenuate neurodegeneration, while also improving systemic metabolic parameters such as glucose regulation and lipid utilization [11–13]. These properties suggest that KD may counteract some of the central and peripheral consequences of hypothyroidism.

While the metabolic and neurobehavioral consequences of hypothyroidism have been well characterized, few studies have explored dietary interventions capable of mitigating these effects. Because hypothyroidism is associated with reduced basal metabolic rate, impaired mitochondrial function, and altered lipid metabolism, KD may represent a potential therapeutic strategy. By promoting fatty acid oxidation and ketone body utilization, KD could partially compensate for impaired glucose-based energy metabolism and may exert neuroprotective and metabolic effects, including improved mitochondrial bioenergetics, reduced oxidative stress, and modulation of neuroinflammatory pathways [8–13]. However, direct evidence linking KD to improved outcomes in hypothyroid models remains limited.

The present study aimed to investigate the effects of KD on behavioral, metabolic, and peripheral tissue parameters in a rat model of PTU-induced hypothyroidism. We evaluated its impact on cognitive and affective behaviors, serum hormone and metabolic markers, and histological changes in the thyroid, liver, white adipose tissue (WAT), and brown adipose tissue (BAT), along with systemic inflammatory markers.

Methods

Animals and Treatment

Adult female Sprague-Dawley rats (300–350 g) were group-housed (3–4 per cage) under controlled conditions (21–23°C, 12:12 h light/dark cycle). Animals were randomly assigned to four groups (n = 7 each): control (C), KD, propylthiouracil (PTU), and KD+PTU.

Hypothyroidism was induced in the PTU and KD+PTU groups by administering 0.05% PTU (BD3813, BLDpharm, Shanghai, China) in drinking water for 6 weeks, as previously described [3, 4]. C and PTU groups received standard chow ad libitum (ARDEN, Türkiye; 20.6% fat, 16.5% protein, 62.9% carbohydrates), whereas KD and KD+PTU groups were switched to a KD (ARDEN, Türkiye; 89.3% fat, 10.2% protein, 0.5% carbohydrates) after the first 2 weeks and maintained on this diet for the remaining 4 weeks, a duration previously shown to induce metabolic and morphological alterations in peripheral tissues [14]. Food and water intake were monitored throughout the study.

Behavioral tests were conducted during the final week in the following order: open field, light/dark box, forced swim, and Morris water maze, with at least 24 h between tests. All assessments were performed by an experimenter blinded to group allocation.

Twenty-four hours after behavioral testing, animals were euthanized. Blood samples were collected via cardiac puncture for glucose and β-hydroxybutyrate (β-HB) measurements. Liver, periovarian WAT, interscapular BAT, and thyroid tissues were harvested for histological and biochemical analyses. Serum was separated by centrifugation (2,000 g, 10 min, 4°C) and stored at −80°C, and tissue samples were fixed in 10% formalin.

Open Field and Light/Dark Box Test

Spontaneous locomotor activity and anxiety-like behavior were assessed using the open field test [12]. Rats were placed in the center of a square arena (100 × 100 cm) and allowed to explore for 5 min. Total distance traveled and time spent in the central zone were recorded using Noldus EthoVision XT. Anxiety-like behavior was further evaluated using the light/dark box test [15]. Rats were placed in the light compartment and allowed to explore for 5 min. Time spent in the light and dark compartments was recorded using the same tracking system.

Forced Swim Test

Depressive-like behavior was assessed using the forced swim test [16]. Rats were individually placed in a transparent cylindrical tank (30 cm diameter, 50 cm water depth) filled with water maintained at 23 ± 1°C for 6 min. Immobility time in the forced swim test, analyzed using EthoVision XT video tracking software, was considered the primary indicator of depressive-like behavior, reflecting behavioral despair.

Morris Water Maze Test

Spatial learning and memory were assessed using the Morris water maze test [12]. The apparatus consisted of a circular pool (100 cm diameter) filled with water (23 ± 1°C), with extra-maze visual cues present. The experiment was conducted over 5 days. During the acquisition phase (days 1–4), rats underwent five trials per day (maximum duration: 60 s) to locate a hidden platform submerged below the water surface. Escape latency was recorded as a measure of spatial learning. Rats that failed to locate the platform were guided to it and allowed to remain for 10 s. On day 5, a probe trial was performed with the platform removed. Rats were allowed to swim for 60 s, and time spent in the target quadrant was recorded. Data were analyzed using Noldus EthoVision XT.

Histological Analysis

Liver, periovarian WAT, interscapular BAT, and thyroid tissues were fixed in 10% neutral-buffered formalin, embedded in paraffin, sectioned at 5 μm, and stained with hematoxylin and eosin. Sections were examined under a light microscope (Olympus BX53, Tokyo, Japan), and representative images were captured at ×200 magnification. For each sample, multiple nonoverlapping fields were randomly selected. Histological evaluation was performed qualitatively by an observer blinded to group allocation.

Enzyme-Linked Immunosorbent Assay

Serum free T4 (fT4) and thyroid-stimulating hormone (TSH) concentrations were measured using enzyme-linked immunosorbent assay (ELISA) kits (BT-LAB, Cat. No. EA0061Ra and Cat. No. E0180Ra). Results are expressed as pmol/L for fT4 and mIU/mL for TSH. Serum levels of tumor necrosis factor alpha (TNF-α) and IL-10 were measured using ELISA kits (BT-LAB, Cat. No. E0764Ra and Cat. No. E0108Ra). Results are expressed as pg/mL. All assays performed according to the manufacturer’s instructions. Absorbance was measured at 450 nm using a microplate reader (Multiskan SkyHigh Microplate Spectrophotometer, Thermo Fisher Scientific Inc.).

Statistical Analyses

Statistical analyses were performed using GraphPad Prism 10. Data are presented as mean ± SEM. Normality was assessed using the Shapiro-Wilk test. Group comparisons were performed using one-way ANOVA, while two-way ANOVA or two-way repeated measures ANOVA was used for time-dependent variables. Tukey’s test was applied for post hoc comparisons. Statistical significance was set at p < 0.05. Sample size was determined a priori using G*Power (effect size f = 0.5, α = 0.05, power = 0.80). No outliers were excluded.

Results

Body Weights and Food Intake

Body weight and food intake data are presented in Figure 1. Two-way repeated measures ANOVA revealed a significant time × group interaction for body weight (F(18, 144) = 2.988, p < 0.001, ηp² = 0.27), whereas main effects of time and group were not significant. Baseline body weights did not differ among groups. On day 42, the PTU group showed significantly lower body weight compared to C (p = 0.017) and KD (p = 0.002), with no other significant pairwise differences.

Fig. 1.

Changes in body weight and food intake across experimental groups. a Body weight measurements over the 6-week experimental period. b Daily food intake from the beginning of the third week – when the ketogenic diet was initiated in the KD and KD+PTU groups – until the end of the study. Body weight data are presented as mean ± SEM. Daily food intake is shown as group total.

For cumulative food intake, two-way ANOVA showed significant effects of time (F(27, 81) = 5.338, p < 0.001, ηp² = 0.64) and group (F(3, 81) = 24.53, p < 0.001, ηp² = 0.48). Food intake was reduced in the PTU group compared to C (p < 0.001), whereas KD increased intake relative to PTU (p < 0.001). The KD+PTU group showed lower intake than C (p = 0.046) but higher intake than PTU (p = 0.001). No difference was observed between C and KD. These results indicate that PTU administration markedly reduced food consumption, while the KD partially mitigated this effect when coadministered with PTU.

Serum fT4, TSH, Glucose, and β-HB Levels

Serum fT4, TSH, glucose, and β-HB levels are presented in Table 1. One-way ANOVA revealed a significant difference in fT4 levels among groups (F(3, 24) = 9.464, p < 0.001, ηp² = 0.54). Both C and KD groups showed higher fT4 levels compared to PTU and KD+PTU groups (all p < 0.05), whereas no differences were observed between C and KD or between PTU and KD+PTU. In contrast, TSH levels did not differ significantly among groups (F(3, 24) = 2.505, p = 0.08, ηp² = 0.24). One-way ANOVA revealed a significant difference in glucose levels among groups (F(3, 24) = 34.94, p < 0.001, ηp² = 0.81). The C group showed higher glucose levels compared to all other groups (all p < 0.001). The KD group had lower glucose levels than PTU (p < 0.001), while no difference was observed between KD and KD+PTU. PTU levels were higher than KD+PTU (p = 0.024), indicating that KD exerted the strongest glucose-lowering effect. Serum β-HB levels also differed significantly among groups (F(3, 24) = 49.66, p < 0.001, ηp² = 0.86). Both KD and KD+PTU groups showed higher β-HB levels compared to C and PTU (all p < 0.001), with no differences between KD and KD+PTU or between C and PTU. These findings confirm successful induction of ketosis by the KD.

Table 1.

Serum fT4, TSH, glucose, and β-HB levels

​	fT4, pmol/L	TSH, mIU/mL	Glucose, mg/dL	β-HB, mmol/L
Control	52.94±2.74	1.55±0.05	91.01±4.16	1.25±0.06
KD	50.19±3.00a	1.68±0.11	55.00±1.66b,c	2.45±0.14b,c
PTU	38.01±3.11b	1.48±0.06	73.00±1.38b	1.22±0.07
KD+PTU	37.25±1.30b	1.42±0.04	61.29±2.57a,b	2.47±0.10b,c

Data are expressed as mean ± SEM. Statistical analysis was performed using one-way ANOVA followed by Tukey’s post hoc test.

β-HB, β-hydroxybutyrate.

^a p < 0.05 vs. PTU.

^b p < 0.01 vs. control.

^c p < 0.01 vs. PTU.

Open Field and Light/Dark Box Test Results

In the open field test, no significant differences were observed among groups in total distance traveled, average velocity, or time spent in the central zone (all p > 0.05). Similarly, in the light/dark box test, time spent in the light compartment did not differ among groups (F(3, 24) = 0.1663, p = 0.92, ηp² = 0.02). These findings indicate that locomotor activity and anxiety-like behavior were not affected by the experimental conditions.

Forced Swim Test Results

Forced swim test results are presented in Figure 2. A significant group effect was observed for immobility time (F(3, 24) = 7.292, p = 0.01, ηp² = 0.48). The PTU and KD+PTU groups showed higher immobility compared to C and KD groups (all p < 0.05), with no differences between PTU and KD+PTU or between C and KD. These findings indicate that both hypothyroidism and the combination of KD with hypothyroidism increased depressive-like behavior, as reflected by longer immobility in the FST, whereas C and KD animals showed similarly lower immobility.

Fig. 2.

Immobility durations in the forced swim test. Data are presented as mean ± SEM. *p < 0.05; **p < 0.01, mean values significantly different from the control and KD groups.

Morris Water Maze Test Results

Morris water maze results are presented in Figure 3. Two-way repeated measures ANOVA showed a significant effect of time (F(2.432, 58.37) = 67.51, p < 0.001, ηp² = 0.74), group (F(3, 24) = 8.520, p < 0.001, ηp² = 0.52), and a time × group interaction (F(9, 72) = 3.140, p = 0.003, ηp² = 0.28), indicating that learning performance differed among groups over time. On day 3, the PTU group displaying higher escape latencies compared the C (p = 0.048) and KD groups (p = 0.037). On day 4, multiple significant differences emerged. The PTU group exhibited significantly higher escape latency compared to the C group (p < 0.001) and the KD group (p = 0.001). Similarly, the KD+PTU group showed longer escape latency than the C group (p = 0.017). The KD group also demonstrated slightly higher latency than the C group (p = 0.047). In contrast, there was no significant difference between PTU and KD+PTU groups (p = 0.48) and KD and KD+PTU groups (p = 0.11). These findings indicate that the performance deficits became more pronounced on the final training day, particularly in the PTU and KD+PTU groups, suggesting impaired learning acquisition in these conditions. One-way ANOVA revealed no significant differences in performance among the groups (F(3, 24) = 0.7440, p = 0.54, ηp² = 0.09), indicating that none of the treatments significantly affected memory retrieval during the probe test.

Fig. 3.

Morris water maze test results. a Escape latency across training days. b Group differences in escape latency on individual training days. c Time spent in target quadrant in probe test. Data are presented as mean ± SEM. *p < 0.05; **p < 0.001, mean values significantly different from the control group. ^#p < 0.05; ^##p < 0.001, mean values significantly different from the PTU group.

Histological Analysis

Histological findings are presented in Figure 4. Thyroid sections from C and KD groups showed normal follicular architecture, whereas the PTU group exhibited follicular hyperplasia, epithelial hypertrophy, and colloid depletion. Similar alterations were observed in the KD+PTU group. Liver sections from the C and KD groups demonstrated preserved hepatic architecture with well-organized hepatic cords, normal sinusoidal spaces, and no inflammatory infiltration. In contrast, the PTU group showed periportal inflammatory cell infiltration, bile duct proliferation, and mild hepatocellular disorganization. These alterations were less pronounced in the KD+PTU group, with milder inflammatory infiltration and better preservation of hepatocellular organization.

Fig. 4.

Hematoxylin and eosin (H&E) staining of tissues including thyroid (a), liver (b), WAT (c), BAT (d). The arrows indicate portal triad and inflammation. Scale bar, 200 μm.

In WAT, the C group exhibited uniformly sized unilocular adipocytes with minimal stromal infiltration. The KD group showed largely comparable morphology. In contrast, the PTU group demonstrated marked inflammatory infiltration and stromal vascular changes. The KD+PTU group exhibited more homogeneous adipocytes and reduced inflammatory infiltration compared to the PTU group. BAT from the C and KD groups showed a typical multilocular morphology consistent with active thermogenic tissue. In the PTU group, BAT whitening was observed, characterized by enlarged lipid droplets and reduced multilocular structure. The KD+PTU group displayed smaller lipid droplets and better preservation of multilocular architecture compared with the PTU group.

Inflammation Markers

The analysis of serum TNF-α and IL-10 levels revealed no statistically significant differences among the experimental groups (F(3, 16) = 2.305, p = 0.12, ηp² = 0.30 and F(3, 16) = 0.5011, p = 0.69, ηp² = 0.09, respectively). This result indicate that the experimental treatments did not significantly alter systemic TNF-α or IL-10 levels.

Discussion

In this study, we examined the effects of KD in a PTU-induced hypothyroidism model. PTU successfully induced hypothyroidism, as reflected by reduced serum fT4 levels, impaired spatial learning, and increased depressive-like behavior. KD induced nutritional ketosis, reduced blood glucose levels, and partially attenuated histological alterations in the liver and adipose tissues. However, KD did not improve cognitive or affective impairments or thyroid morphology, and inflammatory markers remained unchanged.

The glucose-lowering effect observed in both PTU and KD groups is consistent with impaired hepatic glucose production in hypothyroidism and the metabolic shift toward ketone utilization during ketosis [5, 6, 8, 9]. The absence of further increases in ketone levels in the KD+PTU group suggests that ketogenesis was primarily driven by dietary intervention rather than thyroid status.

Hypothyroidism impairs hippocampal long-term potentiation, promotes neuroinflammation, and increases amyloid-β accumulation, contributing to Alzheimer-like neuropathology [3, 4]. These alterations may underlie the persistence of cognitive and mood impairments despite restoration of peripheral euthyroidism [7, 17]. KD has attracted interest as a potential adjunctive strategy, as β-HB serves both as an alternative neuronal energy substrate and a signaling molecule that enhances stress resistance and reduces oxidative damage [10]. It has also been shown to attenuate neurodegeneration and modulate brain energy metabolism in experimental models [11, 12]. In the present study, PTU-treated animals exhibited impaired learning and increased depressive-like behavior, consistent with previous reports [3, 4]. However, KD did not improve these deficits, suggesting that ketosis-induced metabolic adaptations are insufficient to compensate for the loss of thyroid hormone signaling. Given the critical role of thyroid hormones in synaptic plasticity, neurogenesis, and neurotransmitter regulation [1, 3], their deficiency may limit the neurotrophic and mitochondrial effects of KD [18]. In addition, the duration of KD exposure may have been insufficient to reverse established structural and functional alterations in the brain [14].

Histological analysis revealed marked peripheral tissue alterations. In line with previous PTU models, thyroid tissue showed follicular shrinkage, epithelial hypertrophy, and colloid depletion in the PTU and KD+PTU groups [19], while KD alone did not affect thyroid morphology.

Hypothyroidism is associated with hepatic structural and metabolic alterations, including steatosis driven by impaired lipid metabolism [5], as well as hepatocellular disorganization, inflammatory infiltration, and bile duct proliferation linked to oxidative stress and cholestatic changes [20]. Consistent with these findings, PTU-treated animals in the present study exhibited similar hepatic alterations. KD partially attenuated these changes, as reflected by reduced inflammation and improved hepatocellular organization. This is in line with previous reports demonstrating hepatoprotective effects of KD, including reduced lipid accumulation, inflammation, and oxidative stress, along with improved mitochondrial function [5]. The absence of changes in circulating TNF-α and IL-10 suggests that these effects are mediated primarily through local tissue-specific mechanisms. Further studies are needed to clarify the underlying molecular pathways.

Hypothyroidism exerts distinct effects on adipose tissues, promoting inflammatory changes in WAT and reducing thermogenic activity in BAT through impaired lipid metabolism and disrupted UCP1-mediated pathways [21–25]. In the present study, PTU-treated animals exhibited inflammatory infiltration and stromal alterations in WAT, as well as BAT whitening and loss of multilocular structure, consistent with previous reports [21–25]. KD partially ameliorated these changes, improving adipocyte morphology in WAT and preserving multilocular architecture in BAT. These findings align with studies showing that KD enhances thermogenic programming and modulates adipose tissue metabolism [26–29].

This study has several limitations. The absence of a thyroid hormone replacement group precluded evaluation of potential synergistic effects between KD and standard therapy. The use of only female rats limits generalizability, and the relatively short duration of KD may not capture long-term neurobehavioral outcomes. Planned brain tissue analyses could not be performed due to technical issues, limiting mechanistic interpretation. Finally, although histological changes were identified, the lack of quantitative analyses (e.g., lipid content, mitochondrial function, and protein expression) restricts interpretation of tissue-specific effects.

Conclusion

In this PTU-induced hypothyroidism model, KD induced ketosis, reduced glucose levels, and partially attenuated hepatic and adipose tissue alterations but did not improve cognitive or depressive-like behaviors. These findings suggest that persistent thyroid hormone deficiency limits the neuroprotective effects of KD. Thus, restoration of thyroid hormone signaling appears necessary for improving neurobehavioral outcomes, and combining KD with hormone replacement may represent a promising therapeutic approach. Future studies should further investigate the underlying mechanisms and evaluate combined or alternative metabolic strategies, including exogenous ketone supplementation and potential sex-related differences.

Acknowledgments

The authors would like to thank Dokuz Eylül University, Faculty of Medicine, Department of Physiology Laboratory, for providing the laboratory infrastructure and technical support necessary for this study.

Statement of Ethics

All experimental procedures were conducted in accordance with international guidelines for the care and use of laboratory animals, including the European Parliament and Council Directive 2010/63/EU, and adhered to the principles of Replacement, Reduction, and Refinement. The study protocol was reviewed and approved by the Animal Care and Use Committee of Dokuz Eylul University, Faculty of Medicine (Protocol No: 32/2023; approval date: July 5, 2023). Experimental animals were housed and maintained in the Experimental Research Laboratory of the Department of Physiology, Faculty of Medicine, Dokuz Eylul University, under standard laboratory conditions.

Conflict of Interest Statement

The authors have no conflicts of interest to disclose.

Funding Sources

This study was not supported by any sponsor or funder.

Author Contributions

Erhan Caner Akkaya participated in the experimental procedures and biochemical analysis, performed statistical analyses, and wrote the manuscript. Rabia Ilgin and Servet Kızıldağ participated in the experimental procedures and biochemical analysis. Asuman Argon performed histological examination. Ferda Hoşgörler supervised the study. All authors reviewed and approved the final version of the manuscript.

Funding Statement

This study was not supported by any sponsor or funder.

Data Availability Statement

The data that support the findings of this study are available upon request from the corresponding author.