This study investigated the impact of a short-term (5-day) high-caloric diet (HCD), consisting of an additional 1,500 kcal/day of ultra-processed snacks, on brain insulin action in healthy-weight men. The research aimed to address a gap in understanding the developmental trajectory of brain insulin responsiveness, a crucial factor in regulating energy metabolism and feeding behavior. The study employed a nonrandomized controlled design with two groups: an HCD group (n=18, 17 completed) and a control group (n=11) maintaining their regular diet. Participants underwent assessments at baseline, immediately after the intervention (follow-up 1), and one week after resuming a regular diet (follow-up 2).
Brain insulin action was assessed using functional magnetic resonance imaging (fMRI) combined with intranasal insulin (INI) administration. The primary finding was a significant increase in brain insulin activity in specific regions (right insular cortex, left rolandic operculum, and right midbrain/pons) in the HCD group compared to the control group at follow-up 1. However, at follow-up 2, the HCD group showed significantly lower brain insulin activity in different regions (right hippocampus and bilateral fusiform gyrus) compared to the control group. These changes occurred without significant changes in body weight or peripheral insulin sensitivity, suggesting that brain insulin resistance can develop rapidly in response to dietary changes, preceding changes in overall body composition. The HCD also reduced reward sensitivity and increased punishment sensitivity.
The study concludes that brain insulin responsiveness can adapt to short-term dietary changes before noticeable weight gain occurs. This adaptation, specifically the development of brain insulin resistance, may contribute to the development of obesity and associated metabolic diseases. The findings highlight the importance of dietary choices on brain health and metabolic regulation, even in the absence of immediate weight changes. The study's scope is limited to healthy-weight male participants, and the non-randomized design is a potential weakness, although mitigated by the inclusion of a control group and repeated measures.
The study provides compelling evidence for a rapid and dynamic impact of a short-term, high-caloric diet on brain insulin action, independent of changes in body weight or peripheral insulin sensitivity. Crucially, while the study demonstrates associations between dietary changes, brain insulin activity, and reward processing, it cannot definitively establish causation. The observed correlations between brain insulin activity, liver fat, and reward learning are suggestive, but further research is needed to confirm a direct causal link. The study is limited in its ability to make causal claims due to its non-randomized design.
The study's findings have significant practical implications, suggesting that even short-term dietary indiscretions, particularly those involving ultra-processed foods, can have rapid and potentially detrimental effects on brain function. This highlights the importance of dietary choices for maintaining not only metabolic health but also brain health, even in individuals who are not overweight or obese. The observed changes in reward and punishment sensitivity suggest that dietary interventions may need to address not only the nutritional content of food but also its impact on reward processing.
While the study provides valuable insights, it's crucial to acknowledge the limitations. The findings are specific to healthy-weight men, and their applicability to women or individuals with different metabolic profiles remains unknown. The study also did not directly measure brain inflammation, a potential mechanism linking the HCD to changes in brain insulin action and white matter integrity. Despite these uncertainties, the study strongly suggests that dietary interventions aimed at preventing or treating obesity and related metabolic disorders should consider the rapid impact of diet on brain function.
Several critical questions remain unanswered. The study does not determine whether the observed effects are due to the excessive calories, the specific macronutrient composition of the diet, or the ultra-processed nature of the snacks. The long-term consequences of these short-term changes in brain insulin action are also unknown. While the nonrandomized design is a significant limitation, the inclusion of a control group and repeated measures strengthens the internal validity. However, the lack of randomization could introduce bias, and it's possible that unmeasured confounding factors contributed to the observed differences between groups. Future studies should address these limitations to confirm and extend the findings.
The abstract clearly states the main finding: a short-term high-caloric diet (HCD) disrupts brain insulin action in healthy men, and this effect persists even after returning to a regular diet. This is a concise and impactful statement of the core result.
The abstract concisely summarizes the rationale for the study, highlighting the link between brain insulin responsiveness, weight gain, and body fat distribution, and identifying a gap in knowledge regarding the developmental trajectory of brain insulin responsiveness.
The abstract briefly, yet effectively, outlines the study design, including the intervention (5-day HCD), the control group (regular diet), the use of intranasal insulin (INI) and fMRI, and the primary outcome measure (brain insulin activity).
The abstract provides a clear and concise conclusion, stating that brain insulin response can adapt to short-term dietary changes before weight gain, potentially contributing to obesity and related diseases. This emphasizes the preventative implications of the findings.
The abstract correctly limits the scope of the study to healthy-weight male participants, acknowledging sex differences in response to intranasal insulin. This demonstrates methodological rigor and awareness of potential confounding factors.
This high-impact improvement would enhance the abstract's informativeness and impact. While the abstract mentions "calorie-rich sweet and fatty foods" and "ultra-processed snacks," it doesn't specify the degree of caloric increase. Including this quantitative detail strengthens the abstract by providing a clearer picture of the intervention's intensity. This is crucial for the abstract as it is often read independently, and this information is critical for understanding the study's scope.
Implementation: Add the specific caloric increase (1,500 kcal/day) to the description of the HCD. For example, "...short-term overeating with an additional 1,500 kcal per day of calorie-rich..." or "...a 5-day high-caloric diet (HCD), increasing daily intake by 1,500 kcal, that included...".
This medium-impact improvement would strengthen the abstract's completeness. While the abstract mentions "associated diseases," it does not specify which diseases. Briefly listing a few key examples (e.g., type 2 diabetes, metabolic syndrome) would make the statement more concrete and impactful. The abstract should stand alone, and providing these examples enhances its informativeness.
Implementation: Add examples of associated diseases. For example, "...may facilitate the development of obesity and associated diseases, such as type 2 diabetes and metabolic syndrome." or "...obesity and related metabolic diseases."
This low-impact improvement would slightly improve the abstract's clarity. The phrase "outlasted the time-frame of its consumption" is somewhat vague. Replacing it with a more precise phrase would enhance readability. The abstract should be as clear and concise as possible, and this change improves precision.
Implementation: Replace "outlasted the time-frame of its consumption" with a more specific phrase, such as "persisted after returning to a regular diet" or "remained disrupted one week after the diet ended."
The introduction effectively establishes the context by highlighting the detrimental effects of insulin resistance in both the periphery and the central nervous system. It clearly distinguishes between the healthy state, where insulin acts in an anorexigenic fashion, and the insulin-resistant state, where this regulation is disrupted.
The introduction succinctly connects aberrant brain insulin response to negative health outcomes, including higher visceral adipose tissue mass, impaired peripheral metabolism, increased fat mass regain after lifestyle interventions, and the promotion of metabolic, psychiatric, and neurodegenerative diseases. This establishes the significance of the research.
The introduction clearly identifies a critical gap in knowledge: the unclear developmental trajectory of brain insulin responsiveness in humans. This directly justifies the study's aim to investigate the effects of a high-caloric diet on brain insulin action.
The introduction concisely describes the study's objective and methodology, including the use of a high-caloric diet (HCD), a control group, intranasal insulin (INI) application, and functional magnetic resonance imaging (fMRI). This provides a clear overview of the experimental approach.
The introduction explicitly states the study's limitation to male participants due to known sex differences in response to intranasal insulin. This demonstrates methodological rigor and awareness of potential confounding factors.
This high-impact improvement would strengthen the introduction's justification for the study. While the introduction mentions the link between disrupted brain insulin responsiveness and various diseases, it doesn't explicitly state why understanding the early development of this disruption is crucial. Emphasizing the potential for early intervention and prevention would significantly enhance the introduction's impact. The introduction is the place to establish the importance of the research question, and this addition would solidify that importance.
Implementation: Add a sentence or two explicitly stating the importance of understanding the early development of brain insulin resistance. For example: "Understanding the early stages of brain insulin resistance development is crucial for identifying potential intervention points to prevent or delay the onset of obesity and related metabolic disorders." or "Early detection and intervention in brain insulin resistance could have significant implications for preventing the progression to more severe metabolic and neurological conditions."
This medium-impact improvement would enhance the introduction's clarity and specificity. The introduction mentions "ultra-processed snacks" but doesn't provide examples. Giving a few concrete examples (e.g., sugary drinks, packaged pastries, processed meats) would make the intervention more tangible and relatable for the reader. The introduction should be clear and accessible, and this detail adds to that clarity.
Implementation: Include a few examples of the ultra-processed snacks used in the HCD. For example: "...calorie-rich ultra-processed snacks (such as sugary drinks, packaged pastries, and processed meats) in addition to the regular diet..." The specific examples can be chosen to reflect the most common or representative items in the diet.
This medium-impact improvement would strengthen the introduction by providing a more complete picture of the study design. While the introduction mentions a "nonrandomized controlled design," it doesn't explain the rationale for using a nonrandomized design. Briefly explaining the reasoning (e.g., practical constraints, ethical considerations) would add to the methodological transparency. The introduction sets the stage for the methods, and this clarification strengthens that foundation.
Implementation: Add a brief explanation for the nonrandomized design. For example, "Due to [practical constraints/ethical considerations], a nonrandomized controlled design was employed." Or, if participant preference played a role: "Participants were allowed to choose their preferred group (HCD or control), resulting in a nonrandomized controlled design."
The Results section clearly presents the primary finding: the HCD group exhibited significantly higher brain insulin activity in specific regions (right insular cortex, left rolandic operculum, and right midbrain/pons) immediately after the 5-day diet (follow-up 1) compared to the control group, when adjusted for baseline.
The Results section effectively contrasts the follow-up 1 findings with those at follow-up 2 (one week after resuming a regular diet). It clearly states that the HCD group showed significantly lower brain insulin activity in different regions (right hippocampus and bilateral fusiform gyrus) compared to the control group at this later time point.
The Results section appropriately reports the statistical significance of the findings, including the use of family-wise error (FWE) correction for multiple comparisons. This demonstrates methodological rigor.
The Results section effectively links the observed changes in brain insulin activity to other relevant measures, such as liver fat, reward learning, and dietary intake. This provides a more comprehensive picture of the effects of the HCD.
The section appropriately reports negative findings, such as the lack of difference in hypothalamic response to insulin between groups. This demonstrates a balanced and unbiased presentation of the results.
The section clearly describes the changes in reward and punishment sensitivity, noting that the HCD reduced reward sensitivity and increased punishment sensitivity. It also mentions the persistence of this pattern at follow-up 2, albeit with non-significant effects on individual parameters.
The Results section reports findings related to white matter integrity (fractional anisotropy and mean diffusivity), noting significant differences between the HCD and control groups at follow-up 2. This adds another dimension to the observed effects of the HCD.
This high-impact improvement would enhance the clarity and interpretability of the results. While the Results section mentions "adjusted for baseline," it doesn't explicitly state how this adjustment was performed statistically. Specifying the statistical method used (e.g., ANCOVA, difference scores) is crucial for methodological transparency and allows readers to fully understand the analysis. The Results section is where the core findings are presented, and providing this detail is essential for accurate interpretation.
Implementation: Explicitly state the statistical method used for baseline adjustment. For example: "Brain insulin activity was compared between groups at follow-up 1 and follow-up 2, adjusted for baseline values using analysis of covariance (ANCOVA)." or "...using difference scores (follow-up – baseline)."
This medium-impact improvement would enhance the completeness and clarity of the results. While the Results section mentions "whole-brain corrected," it doesn't specify the cluster-defining threshold (CDT) used before the family-wise error (FWE) correction. Providing the CDT (e.g., p < 0.001 uncorrected) is important for understanding the stringency of the analysis and the spatial extent of the significant clusters. This level of detail is standard practice in neuroimaging studies and is necessary for proper interpretation of the results.
Implementation: Specify the cluster-defining threshold (CDT) used before FWE correction. For example, "... (PFWE < 0.05, whole-brain corrected, cluster-defining threshold: p < 0.001 uncorrected)." This information can be included in the main text or in the figure legends.
This medium-impact improvement would enhance the clarity and flow of the Results section. The transition between reporting the primary brain insulin activity findings and the correlation analyses could be smoother. Adding a brief introductory sentence to explain the purpose of the correlation analyses would improve the logical flow and help readers understand the connection between these different analyses. The Results section should present a coherent narrative, and this transition sentence would strengthen that narrative.
Implementation: Add a brief introductory sentence before presenting the correlation analyses. For example: "To further investigate the relationship between brain insulin activity and other metabolic and behavioral changes, we performed correlation analyses." or "We next examined whether changes in brain insulin activity were associated with changes in liver fat, reward learning, and dietary intake."
This low-impact improvement would improve the clarity of reporting. While the text mentions bootstrapping for the reward and punishment sensitivity analysis, it does not specify the number of bootstrap samples used. Including this information enhances transparency and reproducibility.
Implementation: Specify the number of bootstrap samples. For example, '...reduced reward sensitivity (t(27) = -3.6, Pboot < 0.001, 10,000 bootstrap samples)...'
Fig. 2 | Disrupted brain insulin action after short-term overeating with calorie-rich snacks.
Extended Data Fig. 1 | Liver fat content in the control and high-caloric diet group (HCD) at baseline and 5-days after the high-caloric or regular diet intervention (follow-up 1).
Extended Data Fig. 2 | HCD reduced reward sensitivity and increased punishment sensitivity.
Extended Data Table 4 | Changes in brain insulin activity (CBF) from before to after the intervention
Extended Data Table 5 | Correlation between change in brain insulin responsiveness and metabolic and behavioural changes
The Methods section clearly describes the participant inclusion criteria, including age, BMI, health status, and lifestyle factors. This level of detail is crucial for assessing the study's internal validity and the generalizability of the findings.
The section provides a concise overview of the study design, including the intervention (5-day HCD), the control group (regular diet), and the timing of the assessments (baseline, follow-up 1, and follow-up 2). This gives readers a clear understanding of the experimental protocol.
The Methods section specifies the key methodological components, including the use of fMRI, intranasal insulin administration, oGTTs, whole-body MRI, and a reward-learning task. This provides a comprehensive picture of the techniques used to assess the various outcome measures.
The section provides details about the high-caloric diet (HCD), including the caloric increase (1,500 kcal/day) and examples of the snacks provided. This allows readers to understand the nature and intensity of the dietary intervention.
The Methods section describes the procedures for food diary collection and validation, including the use of a validated software and photographs of consumed food. This strengthens the reliability of the dietary intake data.
The section provides detailed information about the MRI acquisition and preprocessing procedures, including scanner specifications, sequences, and data processing steps. This level of detail is essential for reproducibility.
The Methods section clearly outlines the statistical analyses performed, including the primary analysis approach (flexible factorial design in SPM12), the statistical thresholds used, and the correction for multiple comparisons. It also describes the secondary analyses and the software used.
The section describes the questionnaires used to assess various psychological and behavioral factors, such as impulsivity, eating behavior, and mood. This provides context for understanding potential confounding factors or mediating variables.
The Methods section provides a detailed description of the go/no-go learning task and the reinforcement learning model used to analyze the data. This allows readers to understand how reward learning was assessed and quantified.
This medium-impact improvement would enhance the study's methodological transparency. The Methods section states that participants were instructed to walk fewer than 4,000 steps a day, monitored by a Fitbit. However, it doesn't specify how compliance with this instruction was assessed or enforced. Adding this information would strengthen the methods by providing a clearer picture of how physical activity was controlled. The Methods section should detail all procedures relevant to the study's internal validity, and this is a key aspect of controlling for potential confounding variables.
Implementation: Specify how compliance with the step count restriction was assessed and/or enforced. For example: "Participants' step counts were downloaded from their Fitbit devices at each visit to verify compliance with the instruction to walk fewer than 4,000 steps per day. Participants who consistently exceeded this limit were [reminded of the instructions/excluded from the analysis]." Or, if a threshold for acceptable compliance was used: "Compliance was defined as averaging fewer than 4,000 steps per day across the monitoring period."
This medium-impact improvement would enhance the study's reproducibility. While the Methods section mentions that high-caloric snacks were provided based on palatability ratings, it doesn't describe how these ratings were obtained. Providing details about the palatability rating procedure (e.g., the scale used, the specific snacks rated) would allow other researchers to replicate this aspect of the study more accurately. The Methods section should provide sufficient detail for replication, and this is a key element of the intervention.
Implementation: Describe the palatability rating procedure. For example: "Prior to the baseline visit, participants rated the palatability of a range of high-caloric snacks (including Snickers, brownies, chips, etc.) on a scale from 1 (not at all palatable) to 10 (extremely palatable). The snacks with the highest ratings for each participant were then selected for their individualized HCD packages." The specific scale and snacks can be adjusted to match the actual procedure.
This low-impact improvement would slightly enhance the clarity of the Methods section. The section mentions that diffusion-weighted images were acquired to investigate white matter integrity, but it doesn't explicitly state the purpose of this investigation within the context of the study's aims. Adding a brief phrase to clarify the rationale would improve the logical flow. The Methods section should clearly connect each method to the overall research question, and this clarification would strengthen that connection.
Implementation: Add a brief phrase to explain the purpose of the diffusion-weighted imaging. For example: "Diffusion-weighted images were acquired at each visit before INI administration to investigate whether the HCD altered white matter integrity, potentially impacting communication between brain regions involved in reward processing and cognitive control." The specific rationale can be adjusted to match the study's hypotheses.
This low-impact improvement would enhance the completeness of the Methods section. While the section describes the reinforcement learning model, it doesn't fully detail the model comparison process. Specifying how the integrated Bayesian Information Criterion (iBIC) was used to compare models would improve transparency. The Methods section should provide sufficient detail about all data analysis procedures.
Implementation: Add a sentence clarifying the model comparison process. For example: "Models were compared using the group-level integrated BIC (iBIC48), which combines model fit and model complexity across all measurements. Lower iBIC values indicate a better balance between model fit and complexity."
Extended Data Table 2 | State Questionnaires on Brain MRI day in the fasted state
The Discussion effectively summarizes the main findings of the study, highlighting the rapid adaptation of brain insulin responsiveness to a short-term high-caloric diet (HCD) and the persistence of these effects even after returning to a regular diet. This reinforces the core message of the paper.
The Discussion appropriately places the findings within the context of existing literature, citing relevant studies on insulin resistance, reward processing, and brain inflammation. This demonstrates the study's contribution to the broader field of knowledge.
The Discussion connects the observed changes in brain insulin action to specific brain regions (insula, midbrain, hippocampus, fusiform gyrus) and their known functions in reward processing, memory, and food cue responses. This provides a mechanistic interpretation of the findings.
The Discussion acknowledges the potential role of brain inflammation in the observed effects, particularly in relation to the changes in white matter integrity. This demonstrates a comprehensive consideration of the underlying mechanisms.
The Discussion explicitly acknowledges the study's limitations, including the small sample size, the focus on male participants, and the lack of whole-body insulin sensitivity measurements using the gold standard method. This demonstrates methodological transparency and awareness of the study's limitations.
The Discussion proposes future research directions, such as investigating the effects of the HCD in women and exploring the impact of physical inactivity and different macronutrient compositions. This highlights the remaining questions and potential avenues for further investigation.
This medium-impact improvement would strengthen the Discussion by providing a more balanced perspective. While the Discussion mentions the potential role of brain inflammation, it could more explicitly discuss the limitations of drawing conclusions about inflammation based on the available data (no changes in cytokines immediately after the HCD). The Discussion section is where limitations and alternative interpretations are considered, and this addition would strengthen that aspect.
Implementation: Add a sentence or two explicitly acknowledging the limitations of drawing conclusions about brain inflammation. For example: "While the observed changes in white matter integrity are consistent with obesity-associated brain inflammation, the lack of immediate changes in circulating cytokines suggests that other mechanisms may be involved, or that cytokine changes may occur later or be localized to specific brain regions. Further studies with more direct measures of brain inflammation are needed to confirm this hypothesis."
This medium-impact improvement would enhance the Discussion's depth and impact. The Discussion could more explicitly discuss the potential mechanisms by which the HCD might lead to the observed changes in brain insulin action, beyond the general mention of inflammation. Referencing specific pathways or processes (e.g., altered dopamine signaling, changes in synaptic plasticity) would strengthen the mechanistic interpretation. The Discussion section is the place to speculate on underlying mechanisms, and this addition would add depth to the interpretation.
Implementation: Add a paragraph or a few sentences discussing potential mechanisms. For example: "Several potential mechanisms, beyond inflammation, could contribute to the observed changes in brain insulin action. The HCD, rich in sugar and saturated fat, may alter dopamine signaling in reward pathways, leading to the observed changes in reward and punishment sensitivity. Additionally, the diet could induce changes in synaptic plasticity in regions like the hippocampus and fusiform gyrus, impacting memory and food cue responses. Further research is needed to investigate these specific pathways."
This low-impact improvement would slightly improve the Discussion's clarity. The Discussion mentions that "the brain response to insulin adapts to short-term changes in diet before weight gain." While this is generally true, it could be made more precise by specifying which aspects of the brain response adapt (e.g., increased activity in reward regions, decreased activity in cognitive regions). This adds clarity and precision to the concluding statement.
Implementation: Specify which aspects of the brain response adapt. For example: "...we postulate that specific aspects of the brain response to insulin, such as increased activity in reward-related regions and decreased activity in cognitive-related regions, adapt to short-term changes in diet..."