Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Impulsivity and body fat accumulation are linked to cortical and subcortical brain volumes among adolescents and adults


Obesity is associated not only with metabolic and physical health conditions, but with individual variations in cognition and brain health. This study examined the association between body fat (an index of excess weight severity), impulsivity (a vulnerability factor for obesity), and brain structure among adolescents and adults across the body mass index (BMI) spectrum. We used 3D T1 weighted anatomic magnetic resonance imaging scans to map the association between body fat and volumes in regions associated with obesity and impulsivity. Participants were 127 individuals (BMI: 18–40 kg/m2; M = 25.69 ± 5.15), aged 14 to 45 years (M = 24.79 ± 9.60; female = 64). Body fat was measured with bioelectric impendence technology, while impulsivity was measured with the UPPS-P Impulsive Behaviour Scale. Results showed that higher body fat was associated with larger cerebellar white matter, medial orbitofrontal cortex (OFC), and nucleus accumbens volume, although the latter finding was specific to adolescents. The relationship between body fat and medial OFC volume was moderated by impulsivity. Elevated impulsivity was also associated with smaller amygdala and larger frontal pole volumes. Our findings link vulnerability and severity markers of obesity with neuroanatomical measures of frontal, limbic and cerebellar structures, and unravel specific links between body fat and striatal volume in adolescence.


The prevalence of overweight and obesity in adolescents and adults has rapidly increased worldwide1. Excess weight during adolescence predicts higher morbidity and premature mortality in adulthood2. Among adults, excessive body fat accumulation increases the risk of developing chronic health conditions including cardiovascular disease, Type 2 diabetes and dementia3. There is increasing awareness about the negative influence of body fat, not only on metabolic and physical health, but also on brain health4,5.

Excess body fat accumulation is likely associated with poorer brain health via several mechanisms such as neuroinflammation and changes in the phospholipid composition of brain lipids4,6. However, the association between excess body fat and the volume of brain regions relevant to the pathophysiology of obesity is still unclear, as previous studies have either focused on single brain regions (e.g., the hippocampus) or broad morphological measures6,7,8,9,10. In addition, most previous studies have been conducted with adolescents6 or older adults7,8,11. However, it remains unknown whether the relationship between excess body fat and brain health differs across adolescence and young/middle adulthood. Moreover, past studies using body mass index (BMI) as a proxy for adiposity have yielded mixed findings. For example, increased BMI has been linked to both higher and lower volume of the orbitofrontal cortex (OFC), the insula, hippocampus, striatum, and cerebellum12,13,14,15,16,17,18. In addition to BMI, more precise estimates of body fat accumulation are needed given the relevance of such brain regions for neurodevelopment during adolescence and neurodegeneration during mid to late adulthood. Direct measures of actual body fat content obtained through bioelectrical impedance analysis (BIA) may provide a more accurate indicator of body composition than BMI7.

Another aspect that remains unclear regarding the link between body fat and brain volumes across adolescence and adulthood is the influence of personality traits implicated in the vulnerability for overweight and obesity. For example, studies have consistently shown that reward sensitivity, namely, ‘a purposeful drive to obtain rewarding stimuli’19 is elevated in individuals with obesity20, particularly among those with binge eating disorder21. In contrast, the literature on impulsivity, namely, the tendency to act without forethought22, has been more inconsistent. Specifically, different facets of impulsivity have been linked with BMI and brain volume alterations. BMI was positively associated with the facets of positive and negative urgency23 and lack of perseverance20, the latter of which was negatively correlated with grey matter volume of the anterior cingulate cortex and the insula in healthy individuals24. Impulsivity has also been positively related to the volume of the striatum25.

Although no structural imaging studies have examined the link between impulsivity, body fat and brain volumes, functional imaging activation studies have shown that impulsivity modulates the response of the striatum during food valuation and food choice tasks among people with overweight and obesity26,27. Therefore, trait impulsivity may moderate the relationship between adiposity and the structure of brain regions implicated in reward processing and value-based decision-making (e.g., OFC, striatum). Since impulsivity encompasses different components, and excess weight has been specifically linked to emotion-related facets such as positive and negative urgency, it is also plausible to expect associations with the structure of limbic regions, such as the hippocampus and the amygdala18. In addition to the lack of structural imaging studies on facets of impulsivity, additional issues with the “body fat–brain volume literature” to date include the lack of control for confounders, such as age and gender; making it unclear which effects are attributable to adiposity rather than these confounding variables28.

This study aimed to examine the relationship between body composition (i.e., percent body fat), trait impulsivity and regional brain volumes in a large cohort of adolescents and adults with healthy-weight, overweight and obesity. We used a well-validated measure of body fat and a multidimensional measure of impulsivity (the UPPS-P Impulsive Behaviour Scale), and linked them to brain volume measures of regions relevant for overweight and obesity, while controlling for potential confounders (i.e., age, gender, and total intracranial volume). Specifically, based on current neurobiological models of obesity5,29,30,31, we were interested in the cortical and subcortical brain regions ascribed to inhibitory control (i.e., dorsolateral PFC [dlPFC], medial PFC [mPFC]), reward processing (i.e., OFC, ventral striatum, amygdala), habit formation and compulsive eating (i.e., dorsal striatum), and interoception (i.e., the insula). In addition, we were interested in the cerebellum, which is involved in motor and executive functions relevant to obesity29,30,31.

We hypothesised that (i) in both adolescents and adults, % body fat would be negatively associated with regional volumes of the OFC, dlPFC, mPFC, insula, hippocampus, amygdala, cerebellum, and striatum; (ii) impulsivity would be negatively associated with striatal, limbic, and insula volumes; (iii) impulsivity would moderate the relationship between % body fat and regional brain volumes. In addition, we explored if the associations between % body fat and regional brain volumes differed between adolescents and adults.


Sample Characteristics

Table 1 displays the demographic characteristics, body composition and the UPPS-P scores of the whole sample. Regional brain volumes are summarised for the whole sample (Fig. 1) and separately in adolescents and adults (Table 2).

Table 1 Characteristics of the study sample by developmental age group.
Figure 1

Regional brain volumes (mm3) depicted separately for the left and right hemispheres (controlling for total intracranial volume).

Table 2 Regional brain volumes in mm3 (controlling for total intracranial volume) by developmental age group.

Associations between body fat and brain volumes

Partial correlation analyses in the regions of interest

Figure 2 shows the scatterplot graphs of the significant partial correlation results. We found a positive correlation between % body fat and cerebellum white matter volume (left: r = 0.247, p = 0.008, right: r = 0.251, p = 0.007) and medial OFC volume (left: r = 0.238, p = 0.008). No other correlations between % body fat and brain volumes in the regions of interest survived corrections for multiple comparisons.

Analyses by age group (adolescents versus adults): We found a positive correlation between % body fat and left nucleus accumbens (NAcc) volume in adolescents (r = 0.353, p = 0.005), but not adults (r = 0.062, p = 0.634). No other correlations survived corrections for multiple comparisons.

Regression analyses controlling for developmental and sociodemographic factors

Regional brain volumes were included in the regression models if the partial correlational analyses showed that these regions of interest were significantly associated with % body fat. Results showed that, after adjusting for ICV and sociodemographic factors, higher percent body fat was the only significant predictor of larger cerebellum white matter volume (left: β = 0.247, p = 0.011; right: β = 0.207, p = 0.027; Table 3) and left medial OFC volume (β = 0.212, p = 0.014).

Table 3 Summary of multiple regression analyses including sociodemographic variables, % body fat, impulsivity and brain volumes (mm3).

Analyses by age group (adolescents versus adults): Results showed that, after adjusting for ICV and sociodemographic factors, higher % body fat was the only significant predictor of larger left NAcc volume in adolescents (β = 0.400, p = 0.006), while gender was the only significant predictor in adults (β = −0.458, p = 0.007). The latter finding indicates that being male was associated with a smaller left NAcc volume (Table 4).

Table 4 Multiple regression analyses including sociodemographic variables, % body fat and left NAcc volume (mm3) by developmental age group.

Since % body fat may vary between males and females during adolescence, we conducted an additional sensitivity analysis to test if gender influenced the relationship between body fat and NAcc volume in the sub-group of adolescents. We used a hierarchical regression analysis: Age and ICV were entered in the first block of predictors followed by % body fat and gender in the second block. Finally, the interaction (% body fat x gender) was entered in the third block. Results showed that the interaction between % body fat and gender was not a significant predictor of left NAcc volume, (β = −0.143, p = 0.708, 95% CI = −6.43, 4.39). Thus, male and female adolescents did not show distinct correlations between % body fat and NAcc volume.

Associations between impulsivity and brain volumes

Partial correlation analyses in the regions of interest

Figure 2 shows the scatterplot graphs of the significant partial correlation results. We found a positive correlation between negative urgency and right frontal pole volume (r = 0.258, p = 0.006). We also found a negative correlation between positive urgency and amygdala volume (left: r = −0.253, p = 0.007, right: r = −0.279, p = 0.003). No other correlations between impulsivity and brain volumes in the regions of interest survived corrections for multiple comparisons.

Analyses by age group (adolescents versus adults): Partial correlational analyses between impulsivity and regional brain volumes were conducted separately for adolescents and adults. No correlations survived significance corrections for multiple comparisons.

Figure 2

Scatterplot graphs illustrating the partial correlations between regional brain volumes (mm3) and % body fat, positive urgency or negative urgency (controlling for total intracranial volume). The x-axis of each scatterplot graph represents the standardised residuals of % body fat, positive urgency or negative urgency and the y-axis represents the standardised residuals of regional brain volumes. Note: NAcc = nucleus accumbens, OFC = orbitofrontal cortex and WM = white matter.

Regression analyses controlling for developmental and sociodemographic factors

Regional brain volumes were included in the regression models if the partial correlational analyses showed that these regions of interest were significantly associated with impulsivity. Results showed that after adjusting for ICV and sociodemographic factors, higher positive urgency was the only significant predictor of smaller amygdala volume (Left: β = −0.255, p = 0.011; Right: β = −0.267, p = 0.010; Table 3). No other predictors of brain volumes were significant.

Impulsivity as a moderator of the relationship between body fat and brain volumes

The interaction term between % body fat and impulsivity (UPPS-P total score) accounted for a marginally significant proportion of the variance in left mOFC volume (adjusted for ICV), b = −0.95, t(110) = −1.97, p = 0.051. Visual inspection of the interaction plot showed that as % body fat increased, and impulsivity reduced, left mOFC volume increased (Fig. 3). Specifically, at low impulsivity, individuals with higher % body fat had larger left mOFC volumes than those with lower % body fat, b = 20.54, p = 0.002, CI [7.947, 33.136]. Similarly, at average impulsivity, individuals with higher % body fat had larger left mOFC volumes than those with lower % body fat, b = 12.09, p = 0.024, CI 1.634, 22.556]. In contrast, at high impulsivity, left mOFC volume did not significantly differ depending on % body fat, b = 3.65, p = 0.614, CI [−10.661, 17.957]. There were no other significant interactions between % body fat and impulsivity (all p values > 0.30).

Figure 3

Line graph showing the moderating effect of impulsivity (total UPPS-P score) on the relationship between % body fat and left medial orbitofrontal (OFC) volume (mm3) controlling for total intracranial volume. Values for % body fat and total UPPS-P scores are depicted at low (−1SD), average, and high (+1 SD) levels of the predictor variables.


We aimed to map the association between body fat, impulsivity and subcortical regional brain volumes in a cohort of adolescents and adults across the BMI spectrum. Our study led to several key findings. Specifically, higher percent body fat was associated with larger bilateral cerebellum white matter (WM) and left medial orbitofrontal cortex (OFC) volume in the whole sample comprising adults and adolescents, and with larger left NAcc volume among adolescents only. In addition, negative urgency was correlated with larger frontal pole volume, while positive urgency was correlated with smaller amygdala volume. Finally, we found that impulsivity moderated the relationship between percent body fat and left medial OFC volume in the whole sample.

First, we found a correlation between percent body fat and larger cerebellar WM across the whole sample comprising adults and adolescents. Our finding mirrors results from previous research that used BMI as a measure of obesity13 and is in line with past studies showing positive relationships in the temporal and parietal lobes6,32. However, our finding is not consistent with previous studies that found a negative relationship between BMI and WM volume in the basal ganglia and corona radiata33,34. Although the cerebellum has traditionally been ascribed to motor function, emerging evidence suggests that this brain region plays a key role in reward-based learning and executive control, which are compromised in obesity35,36. The discrepant results across previous studies may be due to differences in sample composition. Studies that found negative correlations typically included participants from an older age group or with a wider age range than those studies that found positive correlations, including the current study.

Several mechanisms have been proposed to explain the positive relationship between adiposity and cerebellar WM volume. One possible explanation is that adiposity leads to chronic inflammation, which directly links changes in peripheral health to poorer white matter integrity, and in particular, a higher density of glial cells in the lipid-based myelin sheath37,38. It may also be that increased body fat results in damage to WM, including higher WM hyperintensities as shown on hippocampal volume in clinical populations with obesity39. Future studies should aim to examine the specific mechanism underlying the relationship between larger cerebellar WM volume and adiposity using advanced MRI techniques such as diffusion tensor imaging (DTI) to assess obesity-related changes in the microstructure of WM.

Second, percent body fat was positively correlated with the volume of the medial OFC; a brain region implicated in reward processing of food cues. Our finding is consistent with some previous research that has found a positive relationship with BMI as a proxy for obesity10,14, but not with other studies that have shown a negative relationship32,40. Furthermore, we also found that the positive association between body fat and left medial OFC volume was moderated by impulsivity. Specifically, among individuals with lower impulsivity, those with higher body fat had larger medial OFC volumes than those with lower body fat. In contrast, among individuals with higher impulsivity, body fat did not significantly predict left medial OFC volume. The finding that impulsivity moderates the association between body fat and medial OFC volume may contribute to explaining the discrepant findings reported in previous studies (i.e., positive versus negative associations between adiposity and medial OFC volume). Previous findings support the idea that the medial OFC plays a key role in the pathophysiology of obesity, particularly in the preference and anticipation of reward41. Importantly, we extend upon previous findings by showing that the association between excess body fat and brain regions implicated in obesity differs depending on personality traits (i.e., impulsivity). Thus, future studies examining the association between adiposity and brain volumes should consider the role of trait impulsivity.

Our results also showed a correlation between percent body fat and NAcc volume in adolescents, but not adults. This finding is in line with the role of the NAcc in reward processing, specifically, the experience of reward42. Previous research has shown that NAcc volumes are positively correlated with another measure of obesity, namely, BMI in adults12,14,17, but here we provide the first evidence of this association among adolescents using body fat as an indicator of obesity. A potential explanation for the discrepancy between current and previous findings may be differences in developmental stages, as neuroscience evidence suggests that the adolescent developmental period is linked to rapid maturation of the reward system43. Recent research in children demonstrates that responsivity to food advertisements and higher NAcc volume are associated with genetic risk for obesity and higher percent body fat, which suggests that larger NAcc volume is a vulnerability factor in obesity44. The NAcc also supports the early formation of unhealthy eating behaviours due to a combination of enhanced reward-related sensitivity in the striatum and delayed development of the prefrontal cortex45. The positive link between adiposity and volume in the NAcc may be explained by increased signalling of adipose tissue to the brain12,14. Taken together, the findings suggest that larger NAcc volume in adolescents with increased body fat may contribute to altered reward-related cue processing. Although not the main focus of our paper, our finding that gender predicted striatal volume in adults, in particular, that women have larger left NAcc volume, supports existing research showing that women have larger dorsal striatum (putamen) and dorsolateral prefrontal cortex volumes14. Overall, these findings provide evidence for age and gender-related differences in the structure of brain regions involved in habit learning, executive control, and reward processing.

Positive urgency was negatively associated with amygdala volume, and negative urgency was positively associated with frontal pole volume independent of adiposity and developmental stage. The inverse association between positive urgency and amygdala volume has also been shown in individuals with addiction (e.g., cocaine dependence and pathological gambling)46. The amygdala plays an important role in reward and emotion processing and is a key component of the corticostriatal circuit, comprising the OFC, anterior cingulate cortex, insula and striatum, which is altered in obesity5,41,47,48. Thus, smaller amygdala volumes might predate higher trait impulsivity and reward sensitivity, and confer vulnerability to obesity, but longitudinal studies are needed to test this notion46. The direct association between negative urgency and frontal pole volume - a region implicated in the integration of cognitive and affective information49, is consistent with previous research that found a positive association between frontal pole volume and delay discounting50. Importantly, our finding provides evidence that the frontal pole is involved not only in behavioural impulsivity (i.e., delay discounting), but also trait impulsivity (i.e., negative urgency). Nevertheless, in other studies, higher trait impulsivity has been linked to smaller PFC brain regions (i.e., the frontal gyrus and orbitofrontal cortex)51. However, the latter study included only healthy adolescents, while we included both adults and adolescents with a wide range of adiposity. The discrepant finding could also be explained by the use of different measures of impulsivity, namely, the UPPS-P or the Barratt Impulsiveness Scale.

Interestingly, percent body fat and trait impulsivity were associated with volumes of distinct and overlapping regional brain areas, which indicates that several brain regions are implicated in obesity independent of impulsivity. Similarly, recent neuroimaging research has shown that BMI and impulsivity are differentially associated with regional brain volumes, namely, the amygdala and hippocampus versus the frontal gyrus, respectively52. Alternatively, it has been proposed that eating-specific impulsivity constructs, such as uncontrolled eating, may show overlapping alterations with obesity and impulsivity, but more markedly so with obesity. This idea is supported by consistent findings that BMI is moderately, positively correlated with eating-specific impulsivity as measured by self-reported uncontrolled eating53,54, but weakly associated with general impulsivity52,55. In addition, eating-specific impulsivity (uncontrolled eating) has been shown to mediate the relationship between general impulsivity and obesity (BMI), suggesting that brain volume alterations relate to differences in general impulsivity, which indirectly affect BMI through eating-specific impulsivity52. Future neuroimaging studies should measure both general and eating-specific impulsivity to disentangle their independent and shared contribution to the pathophysiological mechanisms of obesity.

Our findings of observed correlations between percent body fat and regional brain volumes (i.e., cerebellar, medial OFC) extend previous neuroimaging work in several important ways. First, we used percent body fat as a more proximal measure of excess weight compared to BMI and waist-to-hip ratio used in previous work, which allows us to make more sensitive assessments of the pathophysiology of obesity7. Second, we controlled for important confounders such as total ICV, which was not accounted for in numerous previous studies. Third, we carefully examined the role of age by stratifying our analyses using sample ages as cut-offs, and via controlling for age in overall group analyses.

Despite the above strengths, the current study has several limitations. First, the cross-sectional design prevents the determination of whether regional brain alterations precede or follow obesity and impulsivity. Nevertheless, as impulsivity is a trait measure, it likely precedes obesity, which is supported by both preclinical56 and clinical57 research. However, longitudinal designs are required to establish cause and effect relationships between body fat and brain structural alterations. Second, the different MRI data pre-processing steps, analysis methods, toolboxes (and the specific versions) and metrics (i.e., grey matter density, volume, cortical thickness) used between the current study (i.e., Freesurfer version 4.1.0, volumes) and past studies may have contributed to the inconsistent results. One minor limitation of the Freesurfer toolbox is that it only computes WM for the cerebellum and total cortical volume. Third, we used percent body fat as a proximal indicator of body composition, but we did not measure Fat Free Mass (FFM) as it is typically measured using a different technique, namely, dual energy x-ray absorptiometry (DXA) scans. However, FFM is also increased in obesity and has been shown to have differential effects on brain structure to fat mass and percent body fat58. Fourth, we controlled for a number of potential confounders (i.e., age, gender, ICV), except for ‘Years of education’, which reflects distinct underlying constructs in adolescents versus adults. Specifically, in adolescents, education should be mainly predicted by age, while in adults, education should more closely reflect intelligence. Thus, the ‘Years of Education’ variable was removed to run more comparable analyses in adolescents and adults. Finally, this study allowed for an exploratory cross-sectional comparison of adults and adolescents, which revealed emerging effects linked to developmental stage. Nevertheless, longitudinal, repeated-measures studies are also required to determine individual developmental trajectories of body fat, personality and neuroanatomical fluctuations across the lifespan.

In summary, aside from the above limitations, we demonstrated a positive association between accumulation of body fat and the volume of cerebellar, frontal and striatal brain regions; the latter specific to adolescents, and a novel association between impulsivity and frontal and amygdala volumes in a relatively large cohort of community-recruited participants. The current findings are adequately controlled and representative of the body fat–brain health association in the general population.



The sample comprised 127 individuals (50.4% female) aged between 14 and 45 years with a wide range in SES (monthly income: <600€ to >2500€). BMI values spanned the three intervals of the continuum, that is, normal weight (n = 65, 51.2%), overweight (n = 34, 27%), and obese (n = 28, 22%) based on cut-offs from the International Obesity Task Force (IOTF) for adolescents59 and the World Health Organisation (WHO) for adults3. Given the association between developmental stage, body composition, and brain volumes, participants were classified as either adolescents (n = 63, age range: 14–19 years) or adults (n = 64; age range: 25–45 years), based on a standard definition of adolescence (i.e., 11–19 years)60, which resulted in equal numbers of participants in the two groups.

The sample includes participants from the projects BRAINOBE, INTEROBE and NEUROECOBE, conducted between 2010 and 2015, of which we have previously reported findings for functional MRI23,61,62. Participants were recruited from the general community and from specialised services of the University Hospital “Virgen de las Nieves” in Granada, Spain, as well as schools located in the same geographical area (adolescents with normal weight). The inclusion criteria were as follows: (i) aged between 12 and 45 years old; (ii) absence of history or current eating disorder, assessed by the Eating Disorder Inventory (EDI-2); (iii) absence of history or current neurological or psychiatric disorders (including depression and substance-related disorders), as assessed by participant or participant/parent (for adolescents) interviews based on DSM-IV criteria; (iv) no contraindications to MRI (Magnetic Resonance Imaging) scanning (including claustrophobia and implanted ferromagnetic objects); (v) absence of significant abnormalities on MRI; and (vi) absence of history of brain injury involving loss of consciousness for longer than 5 minutes. All participants had normal or corrected-to-normal vision. For participants under 18 years of age, we obtained consent from a parent and/or legal guardian and assent from participants. The study was approved by the Human Research Ethics Committee of the University of Granada (ethics approval number: 1792/10).


Body composition

A Body Composition Analyser (TANITA SC-330 Scale) was used to measure percent body fat (total body fat/total mass) with bioelectrical impedance analysis (BIA). This method has been shown to provide reliable and valid measures of fatness as illustrated by stronger associations between MRI measurements and BIA than between MRI measurements and waist circumference or BMI63.

UPPS-P Impulsive Behaviour Scale

The UPPS-P64 is a 59-item questionnaire designed to measure five distinct personality facets associated with impulsive behaviour: negative urgency, lack of perseverance, lack of premeditation, sensation seeking, and positive urgency. Urgency (26 items) refers to the tendency to experience strong impulses under conditions of negative affect (negative urgency – 12 items) or positive affect (positive urgency – 14 items); (lack of) perseverance (10 items) refers to the individual’s ability to remain focused on a task that may be boring or difficult; (lack of) premeditation (11 items) refers to the tendency to think and reflect on the consequences of an act before engaging in that act; and finally, sensation seeking (12 items) incorporates two aspects: (a) a tendency to enjoy and pursue activities that are exciting and (b) an openness to trying new experiences that may or may not be dangerous. Each item on the UPPS-P is rated on a four-point scale ranging from 1 (strongly agree) to 4 (strongly disagree). Total scores were obtained for each of the five UPPS-P facets, with higher scores indicating higher levels of impulsivity.

MRI Data Acquisition and Processing

Participants were scanned with a 3T Phillips Achieva X-series scanner at Centro Diagnostico Granada in Spain. For each participant, a 3D volume was acquired using a T1-weighted turbo-gradient-echo sequence (3D-TFE) in the sagittal plane, with a 0.94 × 0.94 × 1.0 mm resolution (160 slices, FOV = 240 × 240 mm2, matrix 256 × 256), TR = 8.3 ms, TE = 3.8 ms, and flip angle = 8°). MRI images were transferred to a Linux workstation for pre-processing and regional brain volumes were extracted using the automated FreeSurfer image analysis suite version 4.1.0 ( The automated FreeSurfer pipeline included motion correction65, non-uniform intensity normalisation (N3) at 500 iterations to correct for intensity non-uniformity artefacts (increased from the default number of iterations of 4)66, automated Talairach transformation, removal of non-brain tissue67, and parcellation of neuroanatomical measures, which were extracted for further statistical analysis.

Statistical Analyses

Data was analysed using the Statistical Package for the Social Sciences version 25 (SPSS; Chicago, IL, USA). All variables met assumptions for multiple regression, including linear independence of predictors as assessed by the Durbin-Watson statistic (range = 1.89 to 2.12). We also found adequate levels of collinearity between predictors as assessed by tolerance (range = 0.22 to 0.97) and VIF values (range = 1.02 to 1.34). Homoscedasticity was assessed via visual inspection of scatterplots (residuals versus predicted values). The variance and co-variance matrices were assessed to check whether the predictors were correlated. BMI and % body fat were highly correlated (r = 0.696, p < 0.001). For this reason, we included % body fat, but not BMI, in the regression models as we were interested in the former as it is a more proximal measure of body composition7. Furthermore, we assessed correlations between the continuous socio-demographic variables (i.e., age, SES) and the brain regions of interest. We found that SES was not significantly correlated with any of the brain volumes of interest (all p’s > 0.05). Thus, SES was not included in the regression analyses to increase power.

We restricted our analyses to cortical (i.e., insula, OFC, PFC [medial and dorsolateral]) and subcortical (i.e., striatum, hippocampus, amygdala, and cerebellum) brain regions that have been previously associated with body fat and/or impulsivity. Additionally, white matter estimates were available for the cerebellum and total cortical volume. First, to examine which cortical and subcortical brain regions were associated with body composition and/or impulsivity, we ran partial correlations between brain volumes, % body fat and UPPS-P impulsivity facets (negative urgency, positive urgency, sensation seeking, lack of premeditation, lack of perseverance) controlling for ICV28. Second, we ran hierarchical regression analyses with % body fat and impulsivity as predictors and volumes of the brain regions linked to adiposity and/or impulsivity as dependent variables, controlling for confounders (i.e., ICV in Step 1, and in Step 2, age, developmental age group, and gender). We run these analyses in both the whole sample and each developmental group separately. In the latter case, we did not include age group as a predictor. For partial correlations, the significance threshold was set at p < 0.01 to protect against Type I error due to multiple comparisons. For regression analyses, p < 0.05 was considered significant.

Finally, we conducted four hierarchical multiple regression analyses via the SPSS macro PROCESS68. These analyses tested the hypothesis that impulsivity moderates the association between % body fat and brain volumes for those regions that were significantly associated with body fat in the partial correlations (i.e., bilateral cerebellum WM, left mOFC, and left NAcc). For each analysis, % body fat was a predictor, UPPS-P total score was the moderator, brain volumes were the dependent variables, and ICV was a covariate. The continuous variables (i.e., % body fat, UPPS-P total score and ICV) were centered to avoid high multicollinearity. We created a product term between impulsivity and % body fat to examine the interaction. To estimate the conditional effects, 5,000 bootstrap samples were used.

Compliance with Ethical Standards

Research involving Human Participants and/or Animals: All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. This article does not contain any studies with animals performed by any of the authors. Informed consent was obtained from all individual participants included in the study and/or their legal gaurdians.

Data Availability

The datasets generated during and/or analysed during the current study are available in the figshare repository,


  1. 1.

    Ng, M. et al. Global, regional, and national prevalence of overweight and obesity in children and adults during 1980–2013: a systematic analysis for the Global Burden of Disease Study 2013. The lancet 384, 766–781 (2014).

    Article  Google Scholar 

  2. 2.

    Reilly, J. J. & Kelly, J. Long-term impact of overweight and obesity in childhood and adolescence on morbidity and premature mortality in adulthood: systematic review. Int. J. Obes. 35, 891 (2011).

    CAS  Article  Google Scholar 

  3. 3.

    World Health Organisation [WHO]. Obesity and overweight. Health Available at, (2017).

  4. 4.

    Guillemot-Legris, O. & Muccioli, G. G. Obesity-Induced Neuroinflammation: Beyond the Hypothalamus. Trends Neurosci. (2017).

  5. 5.

    Volkow, N. D., Wise, R. A. & Baler, R. The dopamine motive system: implications for drug and food addiction. Nat. Rev. Neurosci. 18, 741 (2017).

    CAS  Article  Google Scholar 

  6. 6.

    Schwartz, D. H. et al. Adiposity is associated with structural properties of the adolescent brain. NeuroImage 103, 192–201 (2014).

    Article  Google Scholar 

  7. 7.

    Figley, C. R., Asem, J. S., Levenbaum, E. L. & Courtney, S. M. Effects of body mass index and body fat percent on default mode, executive control, and salience network structure and function. Front. Neurosci. 10 (2016).

  8. 8.

    Hsu, F.-C. et al. Adiposity is inversely associated with hippocampal volume in African Americans and European Americans with diabetes. J. Diabetes Complications 30, 1506–1512 (2016).

    Article  Google Scholar 

  9. 9.

    Vainik, U. et al. Neurobehavioural Correlates of Obesity are Largely Heritable. bioRxiv 204917 (2017).

  10. 10.

    Weise, C. M. et al. The obese brain as a heritable phenotype – A combined morphometry and twin study. Int. J. Obes. 2005 41, 458–466 (2017).

    CAS  Google Scholar 

  11. 11.

    Kaur, S. et al. Central Adiposity and Cortical Thickness in Midlife: Psychosom. Med. 77, 671–678 (2015).

    CAS  Google Scholar 

  12. 12.

    Coveleskie, K. et al. Altered functional connectivity within the central reward network in overweight and obese women. Nutr. Diabetes 5, e148 (2015).

    CAS  Article  Google Scholar 

  13. 13.

    Haltia, L. T. et al. Brain White Matter Expansion in Human Obesity and the Recovering Effect of Dieting. J. Clin. Endocrinol. Metab. 92, 3278–3284 (2007).

    CAS  Article  Google Scholar 

  14. 14.

    Horstmann, A. et al. Obesity-Related Differences between Women and Men in Brain Structure and Goal-Directed Behavior. Front. Hum. Neurosci. 5 (2011).

  15. 15.

    Janowitz, D. et al. Association between waist circumference and gray matter volume in 2344 individuals from two adult community-based samples. NeuroImage 122, 149–157 (2015).

    Article  Google Scholar 

  16. 16.

    Kennedy, J. T., Collins, P. F. & Luciana, M. Higher adolescent body mass index is associated with lower regional gray and white matter volumes and lower levels of positive emotionality. Front. Neurosci. 10 (2016).

  17. 17.

    Mole, T. B., Mak, E., Chien, Y. & Voon, V. Dissociated Accumbens and Hippocampal Structural Abnormalities across Obesity and Alcohol Dependence. Int. J. Neuropsychopharmacol. 19 (2016).

  18. 18.

    Moreno-López, L., Soriano-Mas, C., Delgado-Rico, E., Rio-Valle, J. S. & Verdejo-García, A. Brain Structural Correlates of Reward Sensitivity and Impulsivity in Adolescents with Normal and Excess Weight. PLOS ONE 7, e49185 (2012).

    Article  ADS  Google Scholar 

  19. 19.

    Dawe, S. & Loxton, N. J. The role of impulsivity in the development of substance use and eating disorders. Neurosci. Biobehav. Rev. 28, 343–351 (2004).

    Article  Google Scholar 

  20. 20.

    Mobbs, O., Crépin, C., Thiéry, C., Golay, A. & Van der Linden, M. Obesity and the four facets of impulsivity. Patient Educ. Couns. 79, 372–377 (2010).

    Article  Google Scholar 

  21. 21.

    Schag, K., Schönleber, J., Teufel, M., Zipfel, S. & Giel, K. E. Food-related impulsivity in obesity and Binge Eating Disorder - a systematic review: Food-related impulsivity. Obes. Rev. 14, 477–495 (2013).

    CAS  Article  Google Scholar 

  22. 22.

    Dawe, S., Gullo, M. J. & Loxton, N. J. Reward drive and rash impulsiveness as dimensions of impulsivity: Implications for substance misuse. Addict. Behav. 29, 1389–1405 (2004).

    Article  Google Scholar 

  23. 23.

    Delgado-Rico, E., Río-Valle, J. S., González-Jiménez, E., Campoy, C. & Verdejo-García, A. BMI Predicts Emotion-Driven Impulsivity and Cognitive Inflexibility in Adolescents With Excess Weight. Obesity. 20, 1604–1610.

  24. 24.

    Wang, H., Wen, B., Cheng, J. & Li, H. Brain Structural Differences between Normal and Obese Adults and their Links with Lack of Perseverance, Negative Urgency, and Sensation Seeking. Sci. Rep. 7, srep40595 (2017).

    Article  ADS  Google Scholar 

  25. 25.

    Tschernegg, M. et al. Impulsivity relates to striatal gray matter volumes in humans: evidence from a delay discounting paradigm. Front. Hum. Neurosci. 9 (2015).

  26. 26.

    Babbs, R. K. et al. Decreased caudate response to milkshake is associated with higher body mass index and greater impulsivity. Physiol. Behav. 121 (2013).

  27. 27.

    van der Laan, L. N., Barendse, M. E. A., Viergever, M. A. & Smeets, P. A. M. Subtypes of trait impulsivity differentially correlate with neural responses to food choices. Behav. Brain Res. 296, 442–450 (2016).

    Article  Google Scholar 

  28. 28.

    Barnes, J. et al. Head size, age and gender adjustment in MRI studies: A necessary nuisance? Neuroimage 53, 1244–1255 (2010).

  29. 29.

    Augustijn, M. J. C. M. et al. Reduced motor competence in children with obesity is associated with structural differences in the cerebellar peduncles. Brain Imaging Behav. 12, 1000–1010 (2018).

    Article  Google Scholar 

  30. 30.

    Bauer, C. C. C. et al. Child overweight and obesity are associated with reduced executive cognitive performance and brain alterations: a magnetic resonance imaging study in Mexican children. Pediatr. Obes. 10, 196–204 (2015).

    CAS  Article  Google Scholar 

  31. 31.

    Raschpichler, M. et al. Abdominal fat distribution and its relationship to brain changes: the differential effects of age on cerebellar structure and function: a cross-sectional, exploratory study. BMJ Open 3, e001915 (2013).

    Article  Google Scholar 

  32. 32.

    Walther, K., Birdsill, A. C., Glisky, E. L. & Ryan, L. Structural brain differences and cognitive functioning related to body mass index in older females. Hum. Brain Mapp. 31, 1052–1064 (2010).

    Article  Google Scholar 

  33. 33.

    Raji, C. A. et al. Brain Structure and Obesity. Hum. Brain Mapp. 31, 353–364 (2010).

    PubMed  PubMed Central  Google Scholar 

  34. 34.

    Yokum, S., Ng, J. & Stice, E. Relation of Regional Grey and White Matter Volumes to Current BMI and Future Increases in BMI: A Prospective MRI Study. Int. J. Obes. 2005 36, 656–664 (2012).

    CAS  Google Scholar 

  35. 35.

    Miquel, M. et al. Have we been ignoring the elephant in the room? Seven arguments for considering the cerebellum as part of addiction circuitry. Neurosci. Biobehav. Rev. 60, 1–11 (2016).

    Article  Google Scholar 

  36. 36.

    Moreno-López, L. et al. Cocaine use severity and cerebellar gray matter are associated with reversal learning deficits in cocaine-dependent individuals. Addict. Biol. 20, 546–556 (2015).

    Article  Google Scholar 

  37. 37.

    Stillman, C. M., Weinstein, A. M., Marsland, A. L., Gianaros, P. J. & Erickson, K. I. Body–Brain Connections: The Effects of Obesity and Behavioral Interventions on NeurocognitiveAging. Front. Aging Neurosci. 9 (2017).

  38. 38.

    Yokum, S. & Stice, E. Initial body fat gain is related to brain volume changes in adolescents: A repeated-measures voxel-based morphometry study. Obesity 25, 401–407 (2017).

    Article  Google Scholar 

  39. 39.

    Ryan, L. & Walther, K. White matter integrity in older females is altered by increased body fat. Obesity 22, 2039–2046 (2014).

    Article  Google Scholar 

  40. 40.

    Maayan, L., Hoogendoorn, C., Sweat, V. & Convit, A. Disinhibited Eating in Obese Adolescents Is Associated With Orbitofrontal Volume Reductions and Executive Dysfunction. Obesity 19, 1382–1387 (2011).

    Article  Google Scholar 

  41. 41.

    Volkow, N. D., Wang, G.-J. & Baler, R. D. Reward, dopamine and the control of food intake: implications for obesity. Trends Cogn. Sci. 15, 37–46 (2011).

    CAS  Article  Google Scholar 

  42. 42.

    Stice, E. & Yokum, S. Neural Vulnerability Factors that Increase Risk for Future Weight Gain. Psychol. Bull. 142, 447–471 (2016).

    Article  Google Scholar 

  43. 43.

    Ernst, M. The triadic model perspective for the study of adolescent motivated behavior. Brain Cogn. 89, 104–111 (2014).

    Article  Google Scholar 

  44. 44.

    Rapuano, K. M. et al. Genetic risk for obesity predicts nucleus accumbens size and responsivity to real-world food cues. Proc. Natl. Acad. Sci. 114, 160–165 (2017).

    CAS  Article  Google Scholar 

  45. 45.

    Somerville, L. H. & Casey, B. Developmental neurobiology of cognitive control and motivational systems. Curr. Opin. Neurobiol. 20, 236–241 (2010).

    CAS  Article  Google Scholar 

  46. 46.

    Yip, S. W. et al. Gray-matter relationships to diagnostic and transdiagnostic features of drug and behavioral addictions. Addict. Biol. (2017).

  47. 47.

    Clarke, R. E., Verdejo-Garcia, A. & Andrews, Z. B. The role of corticostriatal-hypothalamic neural circuits in feeding behaviour: implications for obesity. J. Neurochem. (2018).

  48. 48.

    Smith, D. G. & Robbins, T. W. The Neurobiological Underpinnings of Obesity and Binge Eating: A Rationale for Adopting the Food Addiction Model. Biol. Psychiatry 73, 804–810 (2013).

    Article  Google Scholar 

  49. 49.

    Ramnani, N. & Owen, A. M. Anterior prefrontal cortex: insights into function from anatomy and neuroimaging. Nat. Rev. Neurosci. 5, 184–194 (2004).

    CAS  Article  Google Scholar 

  50. 50.

    Wang, Q. et al. Dissociated neural substrates underlying impulsive choice and impulsive action. NeuroImage 134, 540–549 (2016).

    Article  Google Scholar 

  51. 51.

    Schilling, C. et al. Cortical thickness of superior frontal cortex predicts impulsiveness and perceptual reasoning in adolescence. Mol. Psychiatry N. Y. 18, 624–30 (2013).

    CAS  Article  Google Scholar 

  52. 52.

    Neseliler, S. et al. Neural and behavioral endophenotypes of obesity. (2018).

  53. 53.

    Price, M., Higgs, S. & Lee, M. Self-reported eating traits: Underlying components of food responsivity and dietary restriction are positively related to BMI. Appetite 95, 203–210 (2015).

    Article  Google Scholar 

  54. 54.

    Vainik, U., Neseliler, S., Konstabel, K., Fellows, L. K. & Dagher, A. Eating traits questionnaires as a continuum of a single concept. Uncontrolled eating. Appetite 90, 229–239 (2015).

    Article  Google Scholar 

  55. 55.

    Meule, A. Commentary: Questionnaire and behavioral task measures of impulsivity are differentially associated with body mass index: a comprehensive meta-analysis. Front. Psychol. 8 (2017).

  56. 56.

    Ferragud, A. et al. High Trait Impulsivity Predicts Food Addiction-Like Behavior in the Rat. Neuropsychopharmacology 39, 2463 (2014).

    Article  Google Scholar 

  57. 57.

    Kishinevsky, F. I. et al. fMRI reactivity on a delay discounting task predicts weight gain in obese women. Appetite 58, 582–592 (2012).

    Article  Google Scholar 

  58. 58.

    Weise, C. M., Thiyyagura, P., Reiman, E. M., Chen, K. & Krakoff, J. Fat-free body mass but not fat mass is associated with reduced gray matter volume of cortical brain regions implicated in autonomic and homeostatic regulation. NeuroImage 64, 712–721 (2013).

    Article  Google Scholar 

  59. 59.

    Cole, T. J. & Lobstein, T. Extended international (IOTF) body mass index cut-offs for thinness, overweight and obesity: Extended international BMI cut-offs. Pediatr. Obes. 7, 284–294 (2012).

    CAS  Article  Google Scholar 

  60. 60.

    Defoe, I. N., Dubas, J. S., Figner, B. & van Aken, M. A. G. A meta-analysis on age differences in risky decision making: Adolescents versus children and adults. Psychol. Bull. 141, 48–84 (2015).

    Article  Google Scholar 

  61. 61.

    Delgado-Rico, E., Soriano-Mas, C., Verdejo-Román, J., Río-Valle, J. S. & Verdejo-García, A. Decreased insular and increased midbrain activations during decision-making under risk in adolescents with excess weight. Obes. Silver Spring. Md 21, 1662–1668 (2013).

    Google Scholar 

  62. 62.

    Verdejo-Román, J., Fornito, A., Soriano-Mas, C., Vilar-López, R. & Verdejo-García, A. Independent functional connectivity networks underpin food and monetary reward sensitivity in excess weight. NeuroImage 146, 293–300 (2017).

    Article  Google Scholar 

  63. 63.

    N. Y. Columbia University & Jikei University. What is visceral fat? In TANITA Body Composition Analyzer BC-420MA 49 (North American Association for the Study of Obesity [NAASO], 2004).

  64. 64.

    Whiteside, S. P. & Lynam, D. R. The five factor model and impulsivity: Using a structural model of personality to understand impulsivity. Personal. Individ. Differ. 30, 669–689 (2001).

    Article  Google Scholar 

  65. 65.

    Reuter, M., Rosas, H. D. & Fischl, B. Highly accurate inverse consistent registration: a robust approach. Neuroimage 53, 1181–1196 (2010).

    Article  Google Scholar 

  66. 66.

    Zheng, W., Chee, M. W. & Zagorodnov, V. Improvement of brain segmentation accuracy by optimizing non-uniformity correction using N3. Neuroimage 48, 73–83 (2009).

    Article  Google Scholar 

  67. 67.

    Ségonne, F. et al. A hybrid approach to the skull stripping problem in MRI. Neuroimage 22, 1060–1075 (2004).

    Article  Google Scholar 

  68. 68.

    Hayes, A. F. Introduction to mediation, moderation, and conditional process analysis A regression-based approach, second edition. (Guilford Publications, 2017).

Download references


This study has been funded by project grants from the Andalusian Health Service (Consejeria de Salud) [PI 0416/2008 (BRAINOBE) and P-10-HUM-6635 (NEUROECOBE)] and the Spanish Ministry of Innovation and Science (MICINN) [PSI2010-17290 (INTEROBE)] awarded to Dr Verdejo-Garcia.

Author information




N.K. analysed the data; N.K. wrote the paper; N.K., V.L., K.C. and A.V.-G. had primary responsibility for final content. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Antonio Verdejo-García.

Ethics declarations

Competing Interests

The authors declare no competing interests.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Kakoschke, N., Lorenzetti, V., Caeyenberghs, K. et al. Impulsivity and body fat accumulation are linked to cortical and subcortical brain volumes among adolescents and adults. Sci Rep 9, 2580 (2019).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing