BACKGROUND:In a recent coordinated meta-analysis of neuroimaging data, we reported gray matter (GM) alterations in acutely underweight patients with anorexia nervosa (AN). Here, we extend these findings by examining individual variation in brain structure within AN, individual-level differentiation between AN and healthy controls (HC), and differences between AN subtypes, with potential relevance for understanding clinical heterogeneity. METHODS AND FINDINGS:We analyzed individual-level data from 11 international sites in the ENIGMA Eating Disorders Working Group, including 570 female participants with AN and 739 HC. We examined cortical thickness, cortical surface area and subcortical volumes in AN versus HC using three complementary approaches: (i) group-level differences in a mega-analysis correcting for age effects, (ii) frequencies of extreme deviations (infra-/supranormal; z < -1.96/z > 1.96) based on normative reference models by the CentileBrain Initiative, and (iii) individual-level classification performance using machine learning. The same analytic framework was applied to compare AN restricting versus binge-eating/purging subtype, additionally correcting for BMI effects. Mega-analyses reinforced previous meta-analytic findings of pronounced and widespread GM deficits in AN compared to HC. Normative modelling revealed that the frequency of infranormal z-scores (23/68 cortical thickness, 13/14 subcortical volume metrics) and supranormal z-scores (35/68 cortical thickness, 17/68 cortical surface area metrics) was significantly higher in AN than expected based on reference data. Individuals with AN could be reliably differentiated from HC using machine-learning classifiers (ROC-AUC = 0.75-0.81). In contrast, neither group-level differences nor frequency of extreme z-scores differed between AN subtypes, and individuals with different subtypes could not be reliably differentiated from each other. Importantly, the observational design cannot distinguish neurobiological differences related to AN from the effects of starvation or low BMI in the AN versus HC analyses. The lack of differences between subtypes does not exclude brain structural differences between AN subtypes that might be detectable with other modalities or analytic approaches. CONCLUSION:Using a mega-analytic approach, we confirm widespread GM deficits in AN, show that these alterations are (in some patients) extreme, and demonstrate that they enable robust classification with superior performance compared to most MRI-based psychiatric classification studies. The absence of differences between AN subtypes may reflect shared neurobiology, though other imaging modalities may reveal distinctions beyond brain structure.

Abstract The mechanism underlying persistent COVID-19-related olfactory dysfunction (OD) remains unknown. There are indications for both peripheral (within the nose) and central (in the brain) mechanisms. Patients can experience a decrease in olfactory ability (quantitative OD) or a distorted odour perception (qualitative OD). The mechanisms might differ between these two symptom presentations. In this (f)MRI study, we investigated structural and functional brain alterations in patients with persistent (> 1 month) quantitative OD (Sniffin’ Sticks score < 30.75, n=21) or parosmia (self-reported, n=27) following SARS-CoV-2 infection, aged 19-60 years and, on average, included 472 days after diagnosis (the COVORTS study). Additionally, a control group of 49 normosmic subjects aged 22-60 years was included. Olfactory bulb volume (OBV) was determined by manual segmentation; whole-brain and regional grey matter density (GMD) of olfactory-related areas was assessed using voxel-based morphometry; and odour-induced brain activation was examined with an olfactory fMRI task. When grouped together, patients with OD had a smaller OBV than controls (84.4 vs 95.3 mm3, p =0.026). The two patient groups did not differ in OBV. ROI analysis of olfactory-related brain regions at an exploratory threshold showed lower GMD in primary and secondary olfactory regions in patients compared to controls, which appears to be mainly driven by quantitative OD. No difference in odour-induced brain activation was found. The observed smaller olfactory bulbs, along with modest findings beyond this primary olfactory structure, provide important insights into potential mechanisms of persistent COVID-19-related OD and highlight the need for longer follow-up to evaluate further neural degradation or recovery.

Abstract We directly compare the influences of impact and responsibility considerations on giving behaviour. In moral philosophy, utilitarianism emphasizes the importance of the former, whereas theories of equity and desert argue for the importance of the latter. Our data are from a television show where an audience of one hundred people divides ten thousand euros among three candidates who face financial difficulties, and from independent raters who evaluated attributes of the candidates and their predicaments. We find that the well-being benefit of donations (‘impact’) outweighs the degree to which the candidate had control over the cause of their situation (‘responsibility’). Giving increases more with impact than it decreases with responsibility, and the contribution of impact to the explanatory power of our regression models is approximately twice that of responsibility. Additionally, our analysis shows no evidence of discrimination based on age, gender, or physical attractiveness.

Olfactory dysfunction affects over 20% of the population. Despite progress in understanding its neural pathophysiology, research remains fragmented. This systematic review synthesizes evidence of brain structural and functional measures, and their association with clinical characteristics (e.g., etiology, duration) in patients with olfactory dysfunction. This may help to identify neural correlates and potential neuroimaging biomarkers of olfactory dysfunction's severity and progression. Following PRISMA guidelines, we screened 2374 papers and included 164 studies. Structural MRI studies consistently reported reduced olfactory bulb volume and/or sulcus depth, alongside gray matter reduction in the orbitofrontal cortex, hippocampus, insula, and amygdala in acquired olfactory dysfunction and paradoxical increases in congenital anosmia. Diffusion tensor imaging studies showed widespread white matter abnormalities, with prominent fractional anisotropy reductions. Resting-state and task-based fMRI studies showed heterogeneous, global alterations in connectivity and/or reactivity. PET/SPECT studies generally reported reduced perfusion or hypometabolism in frontal regions, especially in the orbitofrontal regions. Dopamine transporter imaging showed more frequent dopaminergic deficits in Parkinson's and prodromal individuals with hyposmia. Electroencephalography studies, despite methodological heterogeneity, generally found prolonged latencies and reduced amplitudes in olfactory event-related potentials. Across techniques, these brain alterations often showed low-to-moderate correlations with olfactory function. Although etiological and methodological heterogeneity currently obstructs the identification of robust neuroimaging biomarkers of olfactory dysfunction's severity and progression, current evidence indicates that olfactory dysfunction involves widespread structural and functional alterations, mainly in olfaction-related areas, with the orbitofrontal cortex as a key area emerging across techniques. Large-scale, standardized studies are needed to enable stratified diagnosis and personalized prognosis.

Background Drinks with low-no-calorie sweeteners (LNCSs) do not contribute to energy intake but still provide a hedonic experience through sweetness. LNCS can have differential effects on brain areas involved in food intake and reward compared with sugars. Objectives This study determined changes in brain activity and the effect on physiological markers following the ingestion of flavored waters sweetened with the sugar sucrose or LNCS. Methods Thirty healthy individuals participated in a randomized crossover study with 6 treatments. Participants were scanned after an overnight fast using magnetic resonance imaging, including arterial spin labeling to measure cerebral blood flow (CBF), before and after ingestion of 500-mL drinks: water, or equisweet flavored waters with 25 g sucrose, sucralose, stevia extract, allulose+stevia extract, or monk fruit extract. CBF was measured at baseline and 5 and 30 min; gastric content volume was measured at baseline and 25 and 45 min. Serum insulin and glucose were measured, and participants rated their appetite and thirst throughout each visit. Data were analyzed with linear mixed models. Results Primary: Hypothalamus CBF was not differentially affected by any of the drinks. In the ventral tegmental area (midbrain), treatment effects differed, with lower ΔCBF after sucrose than water, sucralose and monk fruit drink ingestion at 30 min (difference: ∼8% ± 3%; all PFDR < 0.05). Exploratory whole-brain analyses showed increased CBF after consumption for allulose+stevia (amygdala) and stevia (putamen) compared with sucrose (PFWE < 0.05). Despite its low energy content, allulose+stevia delayed gastric emptying similar to sucrose, whereas only sucrose increased glucose and insulin concentrations. Conclusions Although flavored waters with LNCS mostly elicit similar neural and gastrointestinal responses as water, they have some distinct effects on the brain compared with 25 g of sucrose, particularly in reward-related brain areas. Further exploration of the neural and physiological effects of allulose and stevia and dose-dependent investigations are warranted.This study was registered at clinicaltrials.gov as NCT05575687 (https://clinicaltrials.gov/study/NCT05575687).

Global lifespan is rising, alongside increasing age-related conditions. Cognitive disorders such as dementia, are projected to triple in number by 2050. Currently, no cure is available for Alzheimer’s disease (AD), the most common form of dementia, and the need for preventative strategies is paramount. Subjective cognitive decline plus (SCD+) lies on the AD continuum, with increased risk of accelerated cognitive decline, providing an ideal target group for early intervention. The microbiota-gut-brain axis (MGBA) is an emerging avenue for prevention, as dietary fibres may beneficially modulate microbiota and microbial metabolites, such as short chain fatty acids and tryptophan-indoles, resulting in improved gastrointestinal and cognitive functioning. PRECODE is a four-armed, randomised, double-blinded, placebo-controlled trial in 164 older adults (60-79 years) with SCD+ and additional “LIfestyle for BRAin Health” (LIBRA) risk factors. The study aims to investigate 26 weeks of supplementation with chicory inulin, resistant dextrin, and seaweed polysaccharide compared to placebo (maltodextrin) on the MGBA. The primary outcome is effects on neurocognition, assessed by brain activation during working memory load measured with blood-oxygen level dependant functional magnetic resonance imaging (BOLD fMRI) and performance during 2-back task. The secondary outcomes include cognition by neuropsychological test battery, brain and intestinal health parameters, and immune and metabolic markers. Findings may guide new preventative modalities in preclinical AD and provide mechanistic insights into the MGBA in individuals at risk for cognitive decline.Clinical trial registrationhttps://clinicaltrials.gov/study/NCT06433037

BACKGROUND:In vitro studies show that goat milk proteins form less compact coagulates in the stomach compared with cow milk proteins, which may facilitate gastric digestion and amino acid (AA) absorption. However, this has not been confirmed in vivo in humans. OBJECTIVES:This study aims to examine gastric digestion and changes in serum AA concentrations after cow milk-derived (cow MC) and goat milk-derived casein (goat MC) ingestion. METHODS:In this single-blind randomized crossover study participants consumed 300 mL of a drink containing 30 g of cow MC or goat MC. Participants underwent gastric magnetic resonance imaging (MRI) scans at baseline and every 10 min ≤60 min postprandially. Blood was drawn at baseline and ≤4 h postprandially. In addition, participants verbally rated their appetite after each MRI measurement. Primary outcomes were gastric emptying and serum AA concentrations. Secondary outcome was gastric coagulation as inferred by image texture metrics. RESULTS:A total of 18 males (age 26 ± 8.3 y, body mass index 23 ± 1.6 kg/m2) completed the study. Gastric emptying half-time was 80 ± 25 min for goat and 85 ± 24 min for cow MC (P = 0.395). In line with this, gastric emptying of the drinks over time was similar [mean difference (MD) 0.77 mL; 95% confidence interval (CI): -6.9, 8.5; P = 0.845]. Serum essential AA (MD -110 μmol/L; 95% CI: -162, -58) was higher over time for cow MC (P < 0.001). The image texture metric contrast was lower for cow MC (MD 0.010; 95% CI: 0.001, 0.020; P = 0.036). CONCLUSIONS:Cow and goat MC have different coagulating properties, as inferred by AA concentrations and supported by image texture analysis. However, overall gastric emptying and the emptying of the liquid and coagulated fractions were similar. This warrants further in vivo research on casein coagulation in the food matrix to help determine the optimal use for cow and goat milk and their protein fractions.

We provide experimental evidence of a positive memory bias that affects individuals' beliefs, decisions to reinvest, and overconfidence in the stock market. Individuals overremember positive investment outcomes of chosen assets and underremember negative ones. Based on their memories, subjects form overly optimistic beliefs about their investment, reinvest too much, and become overconfident about their investment ability relative to others. We further provide evidence on motivation driving the memory bias. This positive memory bias offers a cognitive microfoundation for why gains weight more than losses when people learn from experiences. This helps reconcile various stylized facts in investor beliefs and behavior.

Gastric protein digestion can be influenced by food micro- and macrostructure. Understanding the interplay between microstructure and macrostructural breakdown during oral processing is crucial to enhance the nutritional value of protein-rich foods. This study compared the in vitro gastric protein digestion of whey protein gels differing in microstructure before and after mastication. Whey protein isolate was mixed with kappa-carrageenan to obtain heat-induced gels differing in microstructure (homogeneous, coarse stranded, protein continuous, bicontinuous). Gel boli were collected from 14 participants. The number, size and total surface area of bolus fragments were determined. In vitro gastric protein digestion was quantified following the INFOGEST 2.0 protocol with minor modifications. Before mastication, coarse stranded gels showed the highest digestion rate (1.07 mmol L- 1 center dot g dry matter(- 1)/h), while homogeneous, protein continuous and bi-continuous gels showed lower and similar digestion rates (0.82-0.87 mmol L- 1 center dot g dry matter- 1/h). After mastication, the total surface area of coarse stranded gels increased 7.9-fold leading to a 1.9-fold increase in digestion rate. In contrast, a 3.4-fold increase in total surface area of bi-continuous gels caused a 2.8-fold increase in digestion rate. The total surface area of homogeneous and protein continuous gels increased 3.1- to 3.6-fold upon mastication resulting in a 1.7- to 1.9fold increase in digestion rate. The increase in protein hydrolysis did not correlate with the degree of structural breakdown after mastication across gels differing in microstructure. We conclude that the impact of microstructure of whey protein gels on in vitro gastric protein digestion is sustained after oral structural breakdown by mastication.

Unhealthy food cues are omnipresent and promote overconsumption. Although childhood obesity rates are increasing, there is no strict regulation of the marketing of unhealthy foods towards children. This is problematic since the human brain, especially areas important for cognitive control, continues to develop into the 30s. It is not known in how far the brain response to unhealthy food cues varies with body mass index (BMI) and age. To investigate this, 168 children (10-17 y) and 182 adults (30-67 y) from the European IDEFICS cohort were scanned with the use of fMRI while viewing pictures of healthy and unhealthy foods. Children exhibited lower activation in the right dorsolateral prefrontal cortex (dlPFC) compared to adults when exposed to unhealthy food cues. Across all age groups, individuals with higher BMI demonstrated reduced activation in the middle cingulum in response to unhealthy food stimuli. Lastly, the relation between BMI and brain activation in response to unhealthy compared with healthy food stimuli varied with development: in children, higher BMI was correlated with decreased activation in right anterior insula and right dlPFC, whereas no such relationship was observed in adults. These findings suggest that children with higher BMI may be particularly vulnerable to unhealthy food cues. In this light, the lack of regulation regarding unhealthy food marketing targeted at children is concerning, especially considering the global increase in obesity rates.

Textured vegetable proteins (TVPs) differ in structural properties that may affect the protein digestion of TVP-based meat analogues. This study explored the influence of structural properties on in vitro gastric protein digestion of TVPs and TVP-based patties. Eight TVPs differing in macroscopic surface area and microstructural properties (porosity, wall density, etc.) were used. As references, TVPs were ground into powders to remove their porous structure. In vitro gastric protein digestion was determined following the INFOGEST protocol. Structural properties were correlated with free amino group concentrations at different digestion times. For intact TVPs, surface area (r = 0.78) and mean pore size (r = 0.74) showed significant positive correlations with free amino group concentrations at 5 min of digestion. These correlations persisted for patties (at 30-120 min for surface area, 5-60 min for mean pore size). For TVP powders, wall density was the main structural feature, which was negatively correlated with free amino group concentrations at 5 min (r = -0.78) and 30 min (r = -0.91) of digestion. Wall density was not correlated with protein digestion for intact TVPs and patties. We conclude that pore-related rather than wall-related TVP properties dominated protein digestion of TVPs and patties. Larger macroscopic surface area and larger mean pore size contributed more to accelerating protein hydrolysis during digestion of TVPs and TVP-based patties than smaller wall density. These results suggest that altering TVP microstructure could potentially be a way to enhance protein digestibility of plant-based meat analogues.

Unhealthy food cues are omnipresent and promote overconsumption. Although childhood obesity rates are increasing, there is no strict regulation of the marketing of unhealthy foods towards children. This is problematic since children’s brains, especially areas important for cognitive control, do not mature until their early 20s. It is not known in how far the brain response to unhealthy food cues varies with body mass index and age. To investigate this, 168 children (10-17 y; 71 prepubertal children, 97 pubertal children) and 182 adults (30-67 y) from the European IDEFICS cohort were scanned with fMRI while viewing pictures of healthy and unhealthy foods. Pubertal children exhibited lower activation in the right dorsolateral prefrontal cortex (dlPFC) compared to adults when exposed to unhealthy food cues. Across all age groups, individuals with higher body mass demonstrated reduced activation in the middle cingulum in response to unhealthy food stimuli. Lastly, the relation between body mass index and brain activation in response to unhealthy compared with healthy food stimuli varied with development: in prepubertal children, higher body mass index was correlated with decreased activation in right anterior insula and right dlPFC, whereas no such relationship was observed in pubertal children or adults. These findings suggest that pubertal children and prepubertal children with higher body mass index may be particularly vulnerable to unhealthy food cues. In this light, the lack of regulation regarding unhealthy food marketing targeted at is concerning, especially considering the global increase in obesity rates.

Pea protein is an interesting alternative for animal-based proteins due to its good availability, low cost, and relatively balanced amino acid (AA) profile. Its digestibility may be affected by heat treatment and food texture. Our aim was to study in vivo AA absorption kinetics and gastric behavior of pea protein products differing in heat treatment and texture and compare this with in vitro digestion. We included fourteen males in this randomized crossover trial with three iso-caloric and iso-volumetric treatments: a 420-mL heated drink, 420-mL unheated drink and 105-g heated gel (semi-solid) consumed with 315 mL water, all containing 20 g pea protein. Gastric MRI scans were made until 90 min post-prandial. Blood samples were collected at baseline and up to 5 h. All treatments were tested with an in vitro digestion model (INFOGEST). Heat treatment did not alter AA absorption kinetics and gastric emptying. Time to maximum peak was delayed for the gel treatment (total AAs: 66.9 versus 48.0 min for both drinks, essential AAs: 75.4 versus 50.0 and 46.6 min for the drinks). For the gel treatment initial emptying was faster due to the rapid passage of water. In vitro, the degree of hydrolysis was highest for the unheated drink in the gastric phase and for the gel treatment in the intestinal phase. In conclusion, heat treating pea protein products does not affect digestion. In contrast, texture of pea protein products can be altered to influence the rate of gastric emptying and AA absorption without affecting total AA absorption.

Correction for 'Gastric coagulation and postprandial amino acid absorption of milk is affected by mineral composition: a randomized crossover trial' by Elise J. M. van Eijnatten et al., Food Funct., 2024, 15, 3098-3107, https://doi.org/10.1039/D3FO04063A.

Background: During breastfeeding the macronutrient composition of breastmilk changes gradually from relatively low-fat (foremilk) to relatively high-fat (hindmilk), initially exposing the gastrointestinal tract to a relatively low fat concentration. In contrast, infant formulae (IF) are homogenous. Mild processing and addition of milk fat globule membrane (MFGM) may impact gastric emulsion instability, potentially impacting the phased release of nutrients as observed during breastfeeding. Objective: To assess gastric emulsion stability, gastric emptying, and the postprandial plasma metabolome of an experimental minimally processed IF (EF) with an altered fat-globule interface and a control IF (CF). Methods: Twenty healthy males participated in this double-blind randomized crossover trial. Gastric MRI scans and blood samples were obtained before and after consumption of 600 ml CF or EF over a 2-h period. Outcomes included gastric top layer formation, total gastric volume, and blood parameters (free fatty acids (FFA), insulin, glucose, and nuclear magnetic resonance (NMR-)metabolomics). Results: EF showed an earlier onset (13.4 min, p = 0.017), smaller maximum volume (49.0 ml, p = 0.033), and a shorter time to maximum top layer volume (13.9 min, p = 0.022), but similar AUC (p = 0.915) compared to CF. Total gastric volume did not show a treatment*time effect. Insulin concentrations were lower for EF. FFA and glucose did not differ. EF yielded higher serum concentrations of phospholipid- and cholesterol-related metabolites. Conclusion: An EF displayed faster gastric creaming than a CF, thereby potentially better mimicking the behavior of breastmilk which leads to phased release of nutrients into the intestine. Overall physiological benefits of this difference in gastric behavior remain to be studied further in infants.

Gastric digestion of proteins is influenced by multiple factors including microstructure, mechanical properties and structure breakdown during mastication. The interplay between these factors affects protein digestion but is underexplored. This study aimed to investigate the contribution of microstructure, mechanical properties and macrostructure breakdown on in vitro whey protein gastric digestion. Whey protein isolate (WPI) was mixed with different types of polysaccharides (kappa-carrageenan, iota-carrageenan, pectin) at various concentrations to obtain heator acid-induced gels with distinct microstructures (homogeneous, coarse stranded, protein continuous and bicontinuous) and Young's moduli (E, 19-165 kPa). Structural breakdown during mastication was mimicked crudely by cutting single gel cylinders into several smaller cubes to increase the total surface area by a factor of 2.65. In vitro gastric digestion was measured using the INFOGEST 2.0 protocol with minor modifications. Homogeneous heat-induced WPI/kappa-carrageenan gels showed the highest digestion rate followed by protein continuous, coarse stranded and bi-continuous heat-induced WPI/kappa-carrageenan gels with similar E. A 1.47-fold increase in E decreased the digestion rate of acid-induced homogeneous WPI/iota-carrageenan gels by a factor of 0.22. In contrast, a 1.13-1.83-fold increase in E barely changed the digestion rate of acid-induced protein continuous WPI/pectin gels. A 2.65-fold increase in total surface area increased the digestion rate of all gels by a factor of 1.35-2.54 depending on microstructure and mechanical properties. We conclude that the microstructure of protein gels affects in vitro protein gastric digestion and the impact of Young's modulus on in vitro protein gastric digestion depends strongly on the microstructure of protein gels.

We investigate financial experts’ beliefs about climate risk pricing and analyze how those beliefs influence stock return expectations. In a comprehensive survey, we elicit experts’ beliefs using both structured and open-ended questions. We establish that most experts share the view that climate risks are insufficiently reflected in stock prices, yet they hold heterogeneous beliefs about the source and persistence of the mispricing. Through the analysis of open text responses, we delineate distinct mental models used by financial professionals to interpret and predict the asset pricing implications of climate risks. Differences in experts’ mental models explain variation in return expectations in the short-term (1-year) and long-term (10-year). Furthermore, we document that experts’ political leanings and geography determine the type of mental model they hold. In a last step, we show that one widely held mental model, which is based on second-order beliefs, causally affects experts’ return expectations using an information provision experiment.

2AbstractBackgroundGastric fluid plays a key role in food digestion and drug dissolution, therefore, the amount of gastric fluid present in a fasted state may influence subsequent digestion and drug delivery. We aimed to describe intra- and interindividual variation in fasted gastric content volume (FGCV) and to determine the association with age, sex, and body size characteristics.MethodsData from 24 MRI studies measuring FGCV in healthy, mostly young individuals after an overnight fast were pooled. Analysis included 366 participants with a total of 870 measurements. Linear mixed model analysis was performed to calculate intra- and interindividual variability and to assess the effects of age, sex, weight, height, weight*height as a proxy for body size, and body mass index (BMI).ResultsFGCV ranged from 0 to 156 mL, with a mean (± SD) value of 33 ± 25 mL. The overall coefficient of variation within the study population was 75.6%, interindividual SD was 15 mL, and the intraindividual SD was 19 mL. Age, weight, height, weight*height, and BMI had no effect on FGCV. Women had lower volumes compared to men (MD: -6 mL), when corrected for the aforementioned factors.ConclusionFGCV is highly variable, with higher intraindividual compared to interindividual variability, indicating that FGCV is subject to day-to-day and within-day variation and is not a stable personal characteristic. This highlights the importance of considering FGCV when studying digestion and drug dissolution. Exact implications remain to be studied.GRAPHICAL ABSTRACTKEY POINTSFasted gastric content volume is highly variable, both within an individual and between individuals, and should range between 0 and 138 mL in healthy young individuals.Women have lower fasted gastric content volume compared to men; age, body weight and body size were not associated with differences in fasted gastric content volume.Fasted gastric content volume can impact both digestion and drug dissolution, although exact implications of the observed variations remain to be studied.

E-cigarettes are harmful, addictive, and popular. In e-cigarettes, nicotine is often paired with food-flavors. How this pairing of nicotine and food cues influences neural processing warrants investigation, as in smokers, both types of cues activate similar brain regions. Additionally, while most e-cigarettes are sweet, savory e-cigarettes are seemingly absent, although savory flavors are commonly liked in food. To understand how smoking status and type of flavor modulate reactions to food-flavored e-cigarettes, in comparison to actual food, neural and subjective responses to food odors were measured in a 2 (sweet vs. savory odor) x2 (food vs. e-cigarette context) x2 (smokers vs. non-smokers) design in 22 occasional/light smokers and 25 non-smokers. During fMRI scanning, participants were exposed to sweet and savory odors and pictures creating the two contexts. Liking and wanting were repeatedly measured on a 100-unit visual-analogue-scale. Results show that sweet e-cigarettes were liked (Δ = 14.2 ± 1.7) and wanted (Δ = 39.5 ± 3.1) more than savory e-cigarettes, and their cues activated the anterior cingulate more (cluster-level qFDR = 0.003). Further, we observed context-dependent variations in insula response to odors (cluster-level qFDR = 0.023, and = 0.030). Savory odors in an e-cigarette context were wanted less than the same odors in a food-context (Δ = 32.8 ± 3.1). Smokers and non-smokers reacted similarly to flavored product cues. Our results indicate that the principles of flavor preference in food cannot directly be applied to e-cigarettes and that it is challenging to design sweet and savory e-cigarettes to appeal to smokers only.