Meta-analysis finds distinct neural activation patterns to food and monetary rewards in overweight/obesity

This peer-reviewed meta-analysis synthesized evidence from 26 functional magnetic resonance imaging (fMRI) studies to delineate common and distinct neural abnormalities during the processing of food or monetary reward cues in individuals with overweight or obesity (OW/OB). The analysis included a total of 1065 participants from these studies, comprising individuals with OW/OB compared to normal-weight (NW) controls. The studies employed various fMRI tasks where participants were exposed to visual or other cues signaling potential food or monetary rewards. The comparator group was consistently NW individuals undergoing the same experimental paradigms. The meta-analysis used Activation Likelihood Estimation (ALE) to identify consistent spatial patterns of neural activation across the included studies.

The primary outcome was neural activation patterns during reward cue processing. The meta-analysis revealed an overlapping neural signature: individuals with OW/OB showed reduced activation in the left posterior cingulate cortex (PCC) and the insula in response to both food and monetary reward cues compared to NW controls. Specific effect sizes, absolute participant numbers for these regions, and p-values or confidence intervals were not reported in the provided data. The analysis also identified reward-type-specific patterns. In the left middle frontal gyrus (MFG), individuals with OW/OB exhibited decreased activation specifically in response to food reward cues but showed increased activation in response to monetary reward cues.

For key secondary outcomes focusing on food-reward tasks, individuals with OW/OB demonstrated increased neural activation in the bilateral caudate nucleus, hippocampus, anterior cingulate cortex (ACC), and medial prefrontal cortex. Concurrently, they showed decreased activation in the amygdala during these tasks. For monetary-reward tasks, the pattern included increased activation in the right lateral nucleus and hypothalamus, alongside decreased activation in the right subthalamic nucleus (STN) and posterior ventral lateral nucleus. As with the primary findings, specific numerical data on effect sizes and statistical precision for these regional activations were not reported.

Safety and tolerability findings are not applicable to this neuroimaging meta-analysis, as it synthesized data on brain activation patterns rather than an interventional drug or device. The analysis did not report on adverse events, serious adverse events, or discontinuations related to the fMRI procedures themselves.

These results contribute to a growing body of neuroimaging literature attempting to map the neural correlates of obesity. Prior landmark studies and meta-analyses have often focused solely on food reward, reporting hyperactivation in striatal regions (like the caudate) and prefrontal areas. This analysis extends that work by directly comparing neural responses to two different reward types within the same synthesized framework, highlighting both shared deficits (in PCC/insula) and divergent responses (in MFG). It suggests the neural phenotype of obesity may involve a generalized reward processing alteration alongside domain-specific sensitivities.

Key methodological limitations must be considered. The most significant limitation is that all synthesized studies were observational and cross-sectional in design, showing associations rather than establishing causation. It cannot be determined whether the observed neural patterns are a cause, a consequence, or a mere correlate of obesity. The specific tasks, stimuli, and analysis thresholds varied across the 26 included studies, which can introduce heterogeneity. Furthermore, the lack of reported effect sizes, confidence intervals, and p-values for the regional findings limits the ability to assess the precision and strength of the evidence. Participant characteristics like age, sex distribution, and specific BMI ranges were not detailed, which affects generalizability.

The clinical implications are currently theoretical. The findings underscore that obesity is associated with complex, reward-type-specific alterations in brain circuits involved in valuation, motivation, and inhibitory control. For practice, this reinforces the concept that obesity involves neurobiological components, but it does not yet translate to specific diagnostic tests or neuromodulation targets. The suggestion for "reward-type-specific interventions" remains hypothetical; no such interventions were tested in this analysis. Clinicians should be aware of this neuroscience context but continue to rely on established, multifactorial approaches to obesity management.

Several critical questions remain unanswered. The causal direction of the observed associations is unknown. It is unclear how these neural patterns relate to individual differences in treatment response, eating behavior, or long-term weight trajectories. The meta-analysis did not address whether these patterns are modifiable through existing interventions like cognitive behavioral therapy, pharmacotherapy, or bariatric surgery. Furthermore, the analysis did not explore potential confounding factors such as psychiatric comorbidities, dietary habits, or genetic predispositions that might influence both brain function and weight status.