ERP correlates of cognitive control and food-related processing in normal weight and severely obese candidates for bariatric surgery: Data gathered using a newly designed Simon task

Highlights

• The effect of a food distractor on cognitive control increases with the desire-to-eat.

• The obese participants showed a larger interference for the food distractors.

• The PF-N1 was sensitive to food distractors in both groups of participants.

• The obese subjects showed delayed P2 and smaller N2 amplitudes for the distractors.

• Both metabolic states and neural abnormalities seem to affect control toward food.

Abstract

Although there have been suggestions that altered cognitive control and food reward sensitivity contribute to overeating in obese individuals, neurophysiological correlates of these mechanisms have been poorly investigated. The current study investigated event-related potentials (ERP) in 24 severely obese and 26 normal weight individuals in fasting condition, using a novel Simon task with food and object distractors.

The study showed that conflict in the Simon task for the food distractor increased with hunger in both groups but was larger with respect to a neutral condition only in the obese individuals. ERP showed higher N1amplitudes in both groups for food distractor, reflecting early food processing. The P2 latency was delayed and the effect of distractors on N2 amplitude was smaller in the obese subjects, reflecting altered neural mechanisms associated with selective attention and cognitive control, all contributing hypothetically to delay response selection of these individuals faced with food distractor.

Introduction

The spreading obesity pandemic represents a challenge to health care systems and means that there is a rapidly growing population of severely obese individuals (i.e. body mass index BMI > 40 kg/m²) who are at higher risk of mortality and medical co-morbidities such as arterial hypertension, type 2 diabetes, obstructive sleep apnea syndrome, non-alcoholic fatty liver, and steatohepatitis (Ogden, Yanovski, Carroll, & Flegal, 2007). It has recently been suggested that an obesogenic environment (i.e., a place where highly caloric foods are readily available) drives the so-called “hedonic” feeding (i.e., feeding in response to pleasure rather than to nutritional needs; (Appelhans, 2009) increasing the risk of overeating and weight gain. Repeated pairing of palatable foods with rewarding outcomes seems to contribute to the development of maladaptive stimulus-response (S-R) associations between those types of foods and eating behaviors, promoting a hypersensitivity of the striatal dopaminergic reward system to food stimuli (for a review see Kenny, 2011). Following the sensitization of S-R associations, food stimuli become particularly salient, automatically capturing attention. Different experimental paradigms and methodologies have uncovered enhanced attentional salience toward food (i.e. food-related attentional bias) in overweight and obese individuals (for a review see Hendrikse et al., 2015).

Growing evidence suggests that adiposity also has an adverse effect on the brain in terms of functional and structural alterations (García-García et al., 2015; Gustafson, Lissner, Bengtsson, Björkelund, & Skoog, 2004; Kullmann, Schweizer, Veit, Fritsche, & Preissl, 2015; Kurth et al., 2013; Marqués-Iturria et al., 2013; Stanek et al., 2011). Cognitive alterations, such as executive dysfunction (e.g. poor decision-making and poor cognitive control, for reviews, Fagundo et al., 2012; Fitzpatrick, Gilbert, & Serpell, 2013; Prickett, Brennan, & Stolwyk, 2015; Spitznagel et al., 2015) have been reported in severely obese individuals. Deficits in cognitive control processes (e.g. response inhibition, interference control; for a review see Braver, 2012) may contribute to dysregulating food consumption and reducing an obese individual’s ability to achieve long-term goals such as successfully losing weight and maintaining weight loss.

Some studies in the literature have described findings showing that there is an imbalance between food-reward sensitivity and cognitive control abilities in obese individuals (Ziauddeen, Alonso-Alonso, Hill, Kelley, & Khan, 2015). Consistent with this view, behavioral studies have reported finding a correlation between BMI values and response inhibition ability during go/no-go tasks designed to study food-related stimuli in normal weight and obese individuals (Houben, Nederkoorn, & Jansen, 2014) after a 2–3 h fast (Price, Lee, & Higgs, 2005). Neuroimaging studies have also been uncovering higher reactivity to food-related stimuli in reward-related brain regions (e.g. the striatum, amygdala and the orbitofrontal cortex) and lower activity in brain regions linked to cognitive control (e.g. the lateral prefrontal cortex) in obese individuals (Brooks, Cedernaes, & Schiöth, 2013; Nummenmaa et al., 2012). An inverse correlation between BMI values and the activity in frontal brain regions was, moreover, described in connection to a go/no-go task with food-related stimuli (Batterink, Yokum, & Stice, 2010; He et al., 2014). Unfortunately, studies comparing behavioral measures of cognitive control in obese and normal weight individuals have produced conflicting results. While some studies have reported reduced inhibitory control (Calvo, Galioto, Gunstad, & Spitznagel, 2014; Chamberlain, Derbyshire, Leppink, & Grant, 2015; Grant, Derbyshire, Leppink, & Chamberlain, 2015; Mole et al., 2015) and interference control in obese individuals, others have not (Bongers et al., 2015; Hendrick, Luo, Zhang, & Li, 2012).

In view of the excellent temporal resolution provided by event-related potentials (ERP), the technique seems particularly useful to investigate selective attention toward food and cognitive control and to expand our knowledge on the neurocognitive correlates of obesity. To date, only a few studies have investigated both cognitive control and reward sensitivity to food stimuli in simply overweight individuals (i.e., with a BMI between 25 and 30 Kg/m²) and in patients with class I obesity (i.e., with a BMI between 30 and 35 Kg/m²) (Carbine et al., 2018; Hume, Howells, Rauch, Kroff, & Lambert, 2015; Nijs, Franken, & Muris, 2010). An elevated BMI has nevertheless been associated with reduced cognitive control and neurocognitive alterations that are particularly relevant in severely obese patients who are candidates for bariatric surgery.

A novel “affective” Simon task utilizing task-irrelevant images of food or objects was designed for the current investigation. The Simon task, a behavioral measure of interference/conflict resolution, was used here to study spatial S-R interference control (i.e., the ability to select the proper response even when other competing responses are present). Despite the fact that the stimuli’s spatial position is irrelevant to accurate performance of the task, the reaction times (RTs) are faster when the stimuli and response positions spatially correspond or match (i.e., corresponding trial) than when they do not (i.e. non-corresponding trial) – this is so-called “Simon effect”. Conflict between a fast direct automatic pathway and a slow, indirect controlled pathway seems to affect response selection during Simon tasks (Lu & Proctor, 1995; Ridderinkhof, 2002; Simon & Rudell, 1967; Umiltá & Nicoletti, 1990). The stimuli’s location seems to automatically activate the spatially corresponding response arising from long-term S-R associations between perceptual and motor processes and linked to genetic factors or to the synaptic consolidation of over-learned S-R associations (Cohen, Dunbar, & McClelland, 1990; Tagliabue, Zorzi, Umiltà, & Bassignani, 2000). A slower indirect (controlled) route, instead, controls goal-directed behavior activating the appropriate response depending on task demands.

It is thought that a dual-route model could explain “addictive-like” behaviors (Bechara, 2005; Evans & Coventry, 2006; Wiers & Stacy, 2006; Wiers, Gladwin, Hofmann, Salemink, & Ridderinkhof, 2013), which depend on an interaction between two pathways of information processing, known as the reflective and impulsive systems (Hofmann, Friese, & Wiers, 2008). In the former, decisions are made in connection to subjective goals and are elicited as a consequence of voluntary decision processes, including executive functions. In the latter, over-learned behavioral repertoires originating from S-R associations stored in long-term memory in close interaction with perceptual stimulus input are activated. As the two pathways interact during response selection and decision-making processes, it can be assumed that enhanced attention toward food-related stimuli is driven by the impulsive system leading to impulsive eating behaviors in obese individuals in whom inhibitory and cognitive control processes in the reflective pathway are weakened.

The current study was designed to utilize ERP to distinguish between the fast automatic/impulsive processes, possibly occurring early after stimulus onset, and the slow/reflective processes, occurring at later stages of processing and linked to deliberate behaviors. Its aim was to compare the neurophysiological correlates of food-related processing and cognitive control as well as eating attitudes and trait impulsivity in severely obese candidates for bariatric surgery (body mass index BMI > 40 kg/m² or a BMI > 35 kg/m² with comorbid conditions) and in normal weight individuals. We expected the images of the distracting food stimuli to interfere with the obese participants’ selective attention and cognitive control processes.

As some studies have suggested that food-related modulations might occur at later stages of information processing (e.g. P2 and P3) in obese/overweight individuals (Hume et al., 2015; Nijs et al., 2010), we expected to find enhanced amplitudes of these components in the presence of food stimuli with a more pronounced effect in obese individuals. We also chose to analyze the N2 and P3 components because of their connection with cognitive control (Folstein & Van Petten, 2008; Kok, Ramautar, De Ruiter, Band, & Ridderinkhof, 2004; Nieuwenhuis, Yeung, Van Den Wildenberg, & Ridderinkhof, 2003; Roche, Garavan, Foxe, & O’Mara, 2005), selective attention and working-memory updating (Polich, 2007). In addition, according to some studies, the P3 is modulated by the S-R interference effect in the Simon task (Leuthold, 2011; Schiff et al., 2014). Independently of food-related processing, we expected obesity-related cognitive dysfunction (Spitznagel et al., 2015) to manifest itself through smaller amplitudes and/or delayed latencies of late ERP components. In other words, we expected to see that differences between normal weight and obese individuals are reflected in modulations in neurophysiological indexes of selective attention, working memory updating (e.g. P2, P3), and executive cognitive control (i.e. N2). In addition, as a recent study reported finding larger N1 amplitude to food cues in hungry non-obese and persons with a history of dieting (Feig et al., 2017), we expected to see a larger N1 in our obese and normal weight participants that should be larger for food distractors with respect to other distractors.