BMI Predicts Emotion-Driven Impulsivity and Cognitive Inflexibility in Adolescents With Excess Weight




Adolescent obesity is increasingly viewed as a brain-related dysfunction, whereby reward-driven urges for pleasurable foods “hijack” response selection systems, such that behavioral control progressively shifts from impulsivity to compulsivity. In this study, we aimed to examine the link between personality factors (sensitivity to reward (SR) and punishment (SP), BMI, and outcome measures of impulsivity vs. flexibility in—otherwise healthy—excessive weight adolescents. Sixty-three adolescents (aged 12–17) classified as obese (n = 26), overweight (n = 16), or normal weight (n = 21) participated in the study. We used psychometric assessments of the SR and SP motivational systems, impulsivity (using the UPPS-P scale), and neurocognitive measures with discriminant validity to dissociate inhibition vs. flexibility deficits (using the process-approach version of the Stroop test). We tested the relative contribution of age, SR/SP, and BMI on estimates of impulsivity and inhibition vs. switching performance using multistep hierarchical regression models. BMI significantly predicted elevations in emotion-driven impulsivity (positive and negative urgency) and inferior flexibility performance in adolescents with excess weight—exceeding the predictive capacity of SR and SP. SR was the main predictor of elevations in sensation seeking and lack of premeditation. These findings demonstrate that increases in BMI are specifically associated with elevations in emotion-driven impulsivity and cognitive inflexibility, supporting a dimensional path in which adolescents with excess weight increase their proneness to overindulge when under strong affective states, and their difficulties to switch or reverse habitual behavioral patterns.


The prevalence of childhood overweight and obesity has been increasing worldwide over the past 30 years, whereas the search for effective interventions is still currently unfulfilling (1). The drastic psychosocial changes shared by all Western societies (e.g., the unrestricted access to food products and the explosion of the food market) have modified the way we perceive food and regulate food intake, being these processes increasingly modulated by external reward cues at the expense of appropriate nutrient sensing (2). In the last years, excessive eating and obesity are increasingly viewed as a brain-related dysfunction, whereby reward-driven urges for pleasurable foods “hijack” context-driven frontal-executive control (3,4). The risk for this imbalance is higher among adolescents, since they stand amidst neurodevelopmental processes essential to fine-tune the links between the reward system and the executive control system (5,6). Neurobiological studies mainly support this notion—adolescents at high-risk for obesity have heightened activation of the reward system in response to receipt of palatable food (7), and obese adolescents have significantly reduced orbitofrontal volumes (8). However, the neuropsychological correlates of excessive weight and its associated neuroadaptations are not yet well specified—despite the fact that cognitive strategies directed to restore these systems are lined up as promising interventions for pediatric obesity (9).

A strand of evidence suggests that excessive weight adolescents have prominent deficits in impulsivity and inhibitory control (10,11). However, the link between excess weight and impulsivity depends on age—high impulsivity is linked to higher body weight in young kids (8–10 years old), whereas in adolescents (12–14 years old) this association is less pronounced and tends to be inversed—low impulsivity associated with higher weight (12). This interaction may reflect age-related maturation of inhibitory processes (13), or a transition between impulsivity and compulsivity in the development of obesity—similar to what is found in addiction (14). Germane to the latter notion, a second strand of evidence indicates that excessive eating is mainly associated with decline of flexibility skills—a cognitive proxy of compulsivity (15). Germane to this notion, neurocognitive studies in children and adolescents have shown that BMI specifically negatively correlates with performance on cognitive switching probes (16,17).

A related unresolved issue is that of the impact of the personality factors of sensitivity to reward (SR) and sensitivity to punishment (SP) on neuropsychological deficits in adolescent overweight and obesity. This is relevant because psychometric studies indicate that SR positively correlates with trait impulsivity, whereas SP positively correlates with compulsivity—with the exception of positive correlations between SP and the impulsivity dimension of urgency (emotion-driven impulsivity (18,19)). Davis (20) demonstrated that SR significantly predicted higher scores on a measure of emotional eating—overeating when under strong negative states, and they proposed a mediational path between SR, negative emotion-driven eating, and BMI. This result is fitting with evidence showing that both SR and SP correlate with negative urgency (the tendency to experience strong impulses under conditions of negative affect) in obese adults (21). Therefore, the association between these personality factors and excessive eating seems well-established. However, further dissociation of the differential role of SR vs. SP, and of their relative impact on inhibitory control vs. inflexibility is needed to achieve a better understanding of how personality impacts neuropsychological profiles in obese adolescents.

In this study, we aimed to examine the link between personality factors (SR and SP), BMI, and outcome measures of impulsivity vs. flexibility in—otherwise healthy—excessive weight adolescents. We conducted a multidimensional assessment of trait impulsivity (using the UPPS-P scale) and neurocognitive assessments with discriminant validity to dissociate inhibition vs. flexibility deficits (using the process-approach version of the Stroop test). According to previous evidence, we predicted that BMI and SR would significantly predict higher scores on emotion-driven impulsivity among adolescents with excess weight.

Methods and Procedures


Sixty-three adolescents (age range 12–17) participated in this study. Participants were classified as obese (n = 26), overweight (n = 16), or normal weight (n = 21) according to the International Obesity Task Force (IOTF) criteria defined by Cole (22). The demographical data and endocrine characteristics of participants—classified as excess weight vs. normal weight—are summarized in Table 1. We recruited them through educational centers and national health services in order to participate in a larger study aimed to test a novel multidisciplinary intervention for adolescent excess weight. To be included, they had to meet the following criteria: (i) age range between 12 and 17 years old, (ii) BMI values falling within the intervals categorized as overweight or obesity according to the IOTF—for excess weight adolescents, or normal weight values, (iii) normal endocrine characteristics—assessed by Complete Blood Count determinations showing within-normal range levels of insulin, basal glucose, triglycerides, high-density lipoprotein cholesterol, and total cholesterol, and (iv) absence of past/current evidence of medical or psychological disorders. The adolescents with normal weight were recruited in the same demographic areas and had to meet the same inclusion criteria—with the exception of (ii). In order to avoid the inclusion of underweight adolescents within this group, we also checked if potential participants fell within normal BMI values according to age- and gender-adjusted Spanish-specific norms (23)—none of the selected participants met criteria for low weight. All participants completed the Millon Adolescent Clinical Inventory (MACI (24)) and the Eating Disorder Inventory 2 (EDI-2 (25)) in order to assess the presence of clinically significant psychopathological traits. Results are presented as Supplementary Table S1 online). In brief, they showed that both groups did not significantly differ on MACI-indexed personality patterns, expressed concerns (EC) or clinical syndromes (CS), with the exceptions of submissive patterns—increased in excess weight adolescents, and peer insecurity—increased in normal weight adolescents. EDI-2 results gave a similar picture—none of the excess weight participants had clinically significant disorders; nonetheless, as a group, they scored significantly higher than their peers on the drive for thinness and body dissatisfaction scales.

Table 1.  Demographical and clinical descriptive data of the normal weight and excess weight groups: sex, SES, age, BMI, and endocrine parameters
inline image


SR and SP. Sensitivity to Punishment and Reward Questionnaire (SPSRQ (26)): The SPSRQ is a 48 yes—no response item questionnaire aimed to measure two neuropsychological systems driving motivated behavior: the behavioral activation system (SR) and the behavioral inhibition system (SP). The total scores from each scale (SP and SR) were obtained for analyses.

BMI. BMI was calculated for each participant as the ratio of weight in kilograms divided by the square of height in meters.

Impulsivity. UPPS-P Scale (19,27): This is a 59-item inventory designed to measure five distinct personality pathways to impulsive behavior: sensation seeking, (lack of) perseverance, (lack of) premeditation, negative urgency, and positive urgency. The first four dimensions were included in the original version of the UPPS scale (27); the fifth dimension has been included based on recent work by Cyders et al. (28) and Smith et al. (29). Sensation seeking (12 items) incorporates two aspects: (i) a tendency to enjoy and pursue activities that are exciting, and (ii) an openness to trying new experiences that may or may not be dangerous; (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, urgency (12 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). Each item on the UPPS is rated on a 4-point scale ranging from 1 (strongly agree) to 4 (strongly disagree). We obtained the total scores of each of these five UPPS-P dimensions for analyses.

Response inhibition and switching. Delis—Kaplan Executive Function System (D-KEFS) Color-Word Interference Test Stroop (CWIT Stroop (30)): This paper and pencil test is based on the Boston process approach (31), which posits that there is a primary function that each test is designed to measure, but also component functions that contribute to performance on a particular task. Rather than a single test of executive control, the CWIT includes a series of four conditions (C) that are administered to determine whether poor performance is because of specific impairment in the component functions of response speed (C1 + C2), response inhibition (C3 − C1), or response switching (C4 − C3). The first condition (C1) presents patches of colors and participants have to name them as quickly and accurately as they can. The second condition (C2) presents the words “red,” “blue,” and “green” printed in black ink and participants are asked to read aloud the words written. The third condition (C3) introduces the inhibition demand: the words “red,” “blue,” and “green” are printed in incongruent colors ink and participants have to name the color and ignore the word. In the fourth condition (C4), the items are similar to condition three but participants have to switch their response between naming the color of the ink and ignoring the word or reading the word (when the item is inside a little box). Based on our study aims, we used as performance indices the normative scores of response inhibition (C3 − C1), or response switching (C4 − C3).

Statistical analyses. We used independent-sample t-tests to examine differences between groups on demographic, personality, clinical (endocrine characteristics), and outcome variables—in these analyses, overweight and obese adolescents were collapsed to form the excess weight group. We examined intercorrelations between these variables by group using Pearson correlation coefficients. To test the main study hypotheses, we conducted three-step level hierarchical multiple regression models to examine the influence of (i) age, (ii) SR and SP personality traits, and (iii) BMI on the outcome measures of impulsivity (UPPS), disinhibition (Stroop inhibition), and inflexibility (Stroop switching). We computed the changes in R2 associated with the inclusion of each of these steps on the prediction model in order to estimate their separate (and aggregated) contribution to prediction of outcome variables.


Group comparisons

Results are presented in Table 2. Excess weight and normal weight adolescents did not significantly differ on sensitivity to punishment, UPPS dimensions of positive and negative urgency, and lack of premeditation and perseverance. Normal weight adolescents showed increased sensitivity to reward (SPSRQ) and elevated sensation seeking (UPPS-P), but effect sizes were medium (Cohen's d = 0.5), and these differences would not survive correction for multiple comparisons (P < 0.006). No differences were found on performance on the Stroop test.

Table 2.  Descriptive scores and group comparisons for measures of sensitivity to punishment/reward (SPSRQ), impulsivity (UPPS-P), and response inhibition and switching (CWIT Stroop)
inline image


Results are displayed in Table 3. Due to the high number of analyses, and to avoid inflated type I error, here we only refer to significant correlations at P < 0.01. In the excessive weight group, SR was positively correlated with negative urgency and positive urgency. SP was only positively correlated with negative urgency. In addition, negative urgency was positively correlated with BMI. In the normal weight group, SR was significantly positively correlated with negative urgency, positive urgency, and lack of perseverance. Endocrine markers did not show significant correlations with personality or neuropsychological indices.

Table 3.  Correlations between age, BMI, endocrine parameters, UPPS-P, SPSRQ, and STROOP variables for each group
inline image

Regression models

Results are presented in Table 4. Age only predicted scores on lack of perseverance. The block including personality traits (SR and SP) significantly predicted scores on the impulsivity dimensions of sensation seeking, lack of premeditation, positive and negative urgency (SR was the main predictor of all models with the exception of the one on lack of premeditation—predicted by both SR and SP), but failed to predict performance on the Stroop test. Inclusion of BMI significantly increased the predictive capacity of age and personality traits on the impulsivity dimensions of positive and negative urgency, and on performance in the Stroop-switching condition. Figure 1 graphically displays the associations between BMI and each of these variables.

Table 4.  Multiple regression models testing the association between age, SPSRQ, and BMI on UPPS-P and STROOP
inline image
Figure 1.

Regression slopes showing significant associations between BMI and (a) positive urgency, (b) negative urgency, and (c) Stroop switching.


Our findings show that BMI positively predicts levels of positive and negative urgency, and negatively predicts Stroop-switching performance in adolescents with overweight and obesity. SR significantly predicts positive and negative urgency, sensation seeking, and lack of premeditation, but fails to predict cognitive performance in the Stroop test. These findings demonstrate that increases in BMI are specifically associated with elevations in emotion-driven impulsivity and cognitive inflexibility, supporting a dimensional path in which adolescents with excess weight increase their proneness to overindulge when under strong affective states, and their difficulties to switch or reverse habitual behavioral patterns.

Excess weight adolescents, compared with their normal weight peers, had very similar psychological concerns, personality characteristics, and executive control performance. The main significant differences—of medium size—emerged for the traits of SR and sensation seeking, but in both cases excess weight individuals had diminished levels of these traits. Therefore, our findings are not supportive of the “hard” conception of adolescent obesity as a dysfunction characterized by hypersensitivity to reward and disrupted inhibitory control (3,14); in fact, the only trait correlated with BMI within the excess weight group was sensitivity to punishment. Conversely, our dimensional approach showed that BMI is significantly associated with increased impulsivity—but only under strong affective states, and poorer cognitive switching. Since sensitivity to punishment and urgency are psychometrically correlated in healthy youths (28), and both have shown significant associations with dysfunctional thought control and compulsivity (18,32), our findings suggest that adolescent obesity is better characterized by negative reinforcing mechanisms triggering habitual behaviors—then difficult to reverse, than by rash-spontaneous impulsivity mechanisms (33). It is interesting to note that this point is straightforward in the case of the link between SP, negative urgency, and compulsivity, but less intuitive to explain the effects of BMI on positive urgency. Taking into account the personality profile of the excess weight group, we believe that this path may be explained in similar terms, because adolescents with excess weight would tend to be constrained (i.e., they are submissive and dependent) until they find themselves in very positive mood, whereby they would feel entitled to overindulge. The overall pattern of increased sensitivity to punishment is consistent to that found in the whole spectrum of eating disorders (34).

The fact that BMI scores predict Stroop-switching but not Stroop inhibition scores is neatly indicative of the notion that flexibility is the main neuropsychological correlate of adolescent excess weight. This finding is in agreement with those of previous studies showing that neurocognitive deficits associated with poor set-shifting and increased perseveration are typical of children and adolescent with excess weight (8,16,17). Nonetheless, there is also evidence of robust deficits in working memory and inhibitory control—taxed with the classic Stroop Color-Word test—in excess weight adolescents having considerably higher BMI levels (mean of 39.86 vs. 29.15 in this study (8)). Altogether, these findings suggest that impaired flexibility is an early cognitive correlate of the disorder (16), which linearly declines as a function of increased adiposity (16,17); however, the deleterious effects of excess weight on neurocognition may progressively extend to other aspects of executive control as the disorder becomes more severe (8). Interestingly, flexibility deficits are also found in other eating disorders, including anorexia and bulimia (35,36), and evidence on the stability of these deficits suggests that it may be a trait or an endophenotype related to eating disorders (36,37,38). Complementarily, the available evidence in childhood and adolescent obesity suggest that increased inflexibility may also reflect a transition from impulsive to compulsive states—impulsivity correlates with BMI in 8- to 10-year-old kids but not in adolescents (12). This notion would actually give support to proposed parallels between obesity and addiction (9,14). Irrespective of their etiology, flexibility deficits may compromise recovery potential in obesity in a number of ways, including inability to reverse habitual feeding patterns or to change unhealthy lifestyles.

Strengths of this study include the careful selection of the excess weight adolescents, who were matched to their normal weight peers in biochemical and psychological indices; this selection allowed us to elegantly test the neuropsychological assumptions without any medical or psychological confounder. In addition, we used specific well-validated indices of inhibition vs. switching skills, which are regarded as neurocognitive probes of impulsive vs. compulsive behavioral patterns (15), providing a novel approach to the notion that the development of obesity holds several parallels with the neuroadaptations that characterize addiction (14). Relevant shortcomings include the relatively small sample size, which is nonetheless sufficient to implement the regression models we run (39), and the lack of naturalistic or psychometric measures of compulsive or binge-like food intake—which may have strengthen the theoretical links we raised. In conclusion, our findings are supportive of a dimensional approach to adolescent obesity, by which increases in weight and adiposity are associated with less ability to control impulses under strong affects and impaired cognitive flexibility.


This study has been funded by grants PI 0416/2008 (BRAINOBE) from the Andalusian Health Service (Consejería de Salud) and PSI2010-17290 (INTEROBE) from the Ministry of Innovation and Science (MICINN).


The authors declared no conflict of interest.

Supplementary Material

Supplementary material is linked to the online version of the paper at http:www.nature.comoby