Neural Response to Pictures of Food after Exercise in Normal-Weight and Obese Women : Medicine & Science in Sports & Exercise

Journal Logo


Neural Response to Pictures of Food after Exercise in Normal-Weight and Obese Women


Author Information
Medicine & Science in Sports & Exercise: October 2012 - Volume 44 - Issue 10 - p 1864-1870
doi: 10.1249/MSS.0b013e31825cade5
  • Free


The high prevalence of obesity among adults is a significant public health concern in the United States (6). Unhealthy eating behaviors and excess energy intake contribute to weight gain and obesity and may be influenced by environmental, behavioral, or physiological factors (2,3,9,27,32,33). Recently, assessment of neural outcomes in response to viewing pictures of food has been used to index attentional response to food (29) (subsequently referred to here as food motivation). Food motivation may be a correlate of appetite and/or a predictor of energy intake. In addition, recent data suggest that neurologically determined food motivation may be different between normal-weight and obese adults (27).

Exercise modestly counteracts the effect of excessive energy intake by increasing energy expenditure (15,19) and may help prevent weight gain and obesity (5). Interestingly, exercise has also been investigated for its influence on appetite and energy intake (7,14). However, the effect of exercise on appetite and energy intake has not been consistent (14,16,22,23), and the effect of exercise on neurologically determined food motivation is not currently known. In addition, many previous studies assessing the effect of exercise on appetite are limited to subjective measures of appetite, whereas food motivation may be determined objectively. Furthermore, there is deficiency of data examining the extent to which normal-weight and obese adults differ in food-seeking behavior after exercise (24).

One objective way to understand the neural bases of food motivation and attention processes is through scalp-recorded event-related potentials (ERP). An ERP reflects the minute changes in brain electrical activity locked to an external stimulus, such as a picture or tone. One advantage of ERPs relative to other measures of neural activity is temporal sensitivity. The temporal resolution of ERPs is in milliseconds; thus, the rapid allocation and distribution of cognitive resources across food-related and nonfood conditions can be measured.

For the purposes of the current investigation, we used an ERP known as the late positive potential (LPP). The LPP is a midline ERP with a maximal superior–posterior positivity that is frequently associated with the level of engagement and processing elicited by emotional stimuli (11). Moreover, LPP amplitudes are thought to reflect increased processing and attention toward motivational or emotionally intense stimuli that continues after stimulus offset (11). Previous studies of the LPP in food-related paradigms suggest increased LPP amplitudes to food pictures relative to pictures of flowers, particularly when an individual is hungry versus satiated (29).

We are unaware of any studies to date that have examined the role of exercise or the potential interaction between exercise and body mass index (BMI) on neural indices of food motivation, such as the LPP. Therefore, the primary purpose of this study was to objectively compare normal-weight and obese women during two separate conditions (nonexercise; exercise) for food motivation, as determined by electrophysiological responses to food pictures. Secondary outcomes for this study included energy intake and physical activity (PA) over 24 h. We hypothesized that BMI group (normal-weight, obese) and exercise would influence food motivation and PA level.


This quasi experimental crossover study used a matched-subject design with the order of the experimental conditions (nonexercise, exercise) counterbalanced. Measured outcomes for this study were electrophysiological indices of food motivation (see below), energy intake, and PA. Participants were followed for 24 h during each condition, and outcomes were assessed identically during both conditions.


After approval by the university’s institutional review board, written consent from 35 healthy women was obtained. Participant characteristics are listed in Table 1. Participants were classified as normal-weight (n = 18, BMI < 25 kg·m−2) or clinically obese (n = 17, BMI ≥ 30 kg·m−2). All women reported being untrained (no vigorous PA >3 d·wk−1 for 20 min per session) but able to walk comfortably for 45 continuous minutes at a moderate-to-vigorous (MV) intensity. Participants were also premenopausal, right-handed, and matched for age and education. Participants were excluded if they had a chronic or metabolic disease; had an orthopedic impairment; were participating in an extreme diet; had food allergies; were previously diagnosed with anorexia, bulimia, or instances of binge eating; reported alcohol or substance abuse within the past year; used tobacco products; were pregnant or lactating; were using antiepileptic medications, reported a history of learning disability; or had a neurological disorder (e.g., traumatic brain injury, seizure disorder, stroke) and/or attention-deficit/hyperactivity disorder.

Participants’ characteristics at baseline.

Experimental conditions

Each participant completed both the nonexercise and the exercise experimental conditions separated by a week’s time. The order of each condition was counterbalanced across participants. Except for the exercise bout, the testing protocol for each condition was identical, including the same time of morning, same day of the week, after at least 7 h of sleep the previous night, after the same relative dietary preload (energy shake) that morning 2 h before testing, after voiding, and not having consumed caffeine that day or performed vigorous intensity exercise during the previous 24 h. Each participant consumed a modest dietary preload (∼10% of estimated daily energy needs) on the morning of each condition to avoid exercise and testing in a fasted state. Energy needs were determined using the Harris–Benedict equation to predict resting metabolic rate multiplied by an activity factor of 1.3 (21).

Nonexercise condition

The nonexercise condition acted as a control condition in which there was no supervised exercise bout. On arrival at the laboratory at ∼8 a.m., each participant began wearing a GT1M accelerometer (ActiGraph, Pensacola, FL); however, it was removed briefly when performing the body composition assessments noted below. Participant height was measured using a standard wall-mounted stadiometer (Seca, Chino, CA), and body weight was assessed using a digital scale (Tanita Corporation, Tokyo, Japan), accurate to the nearest hundredth pound. Subsequently, participants were given a standard one-piece bathing suit and swim cap to wear and were tested for body composition using the BOD POD. The BOD POD has been previously shown to be valid and reliable compared with dual-energy x-ray absorptiometry (1,20).

After the conclusion of these assessments, each participant completed a computerized task (viewing pictures of food and flowers) while electroencephalogram (EEG) was recorded (see methods in the next sections). Participants were then instructed to resume their normal daily routine; however, they were instructed to record all energy intake and to continue to wear the accelerometer until the following morning (see methods in the next sections). There were no additional recommendations or limitations on energy intake or type or amount of PA/exercise.

Exercise condition

The exercise condition was identical with the nonexercise condition except that each participant completed a single MV intensity exercise bout instead of body composition assessments. Like the nonexercise condition, the accelerometer was worn beginning at ∼8 a.m. (before the exercise bout). Each participant completed the exercise on a motor-driven treadmill at 3.8 mph, 0% grade, for 45 consecutive minutes. The intensity and duration of exercise were chosen based on recommendations from the American College of Sports Medicine (4,12) and on our pilot research that this level was feasible for the target population. All participants subsequently completed the same computerized task (while EEG was recorded) as the nonexercise condition and were then released to follow their normal routine, while recording all energy intake and wearing the accelerometer, until the following morning.

Food motivation experimental task

As noted above, immediately after both the body composition testing and the exercise bout (<1 h), participants completed a computerized passive viewing task while electrophysiological neural activity was recorded using EEG. The passive viewing task consisted of three blocks of 80 pictures per block (240 total trials). Each block consisted of 40 pictures of plated meals and 40 pictures of flowers. Food and flower pictures were chosen owing to the similarity of flowers in picture composition to plated foods and because they were used in previous research on neural correlates of food motivation (29). Pictures were matched on contrast, intensity, and brightness levels and were presented for 2000 ms followed by a 500-ms intertrial interval on a 17-inch computer monitor approximately 20 inches from the participant’s head. Pictures were presented in random order. The task was identical (same stimuli, but presented in a random order) for both sessions.

Dimensional views of emotion processing suggest emotion consists of a confluence of multiple contributing dimensions. Two of the most common contributing dimensions in emotion research are the affective valence (ranging from unpleasant to pleasant) and the emotional arousal (ranging from calm to excited) of the stimuli (17). Thus, to determine whether the food or flower pictures were seen as having different levels of valence or arousal, participants rated valence on a 9-point scale from extremely unpleasant to extremely pleasant and arousal on a 9-point scale from not at all arousing to extremely arousing or emotional (see Lang et al. [17] for description).

Electrophysiological data recording and reduction

EEG was recorded from 128 scalp sites using a geodesic sensor net and Electrical Geodesics, Inc. (EGI; Eugene, OR) amplifier system (20K nominal gain, band-pass = 0.10–100 Hz). EEG was initially referenced to the vertex electrode and digitized continuously at 250 Hz with a 24-bit analog-to-digital converter. Impedances were maintained below 50 kΩ as recommended by the manufacturer. Data were average-referenced offline and digitally low-pass-filtered at 30 Hz. Eye movement and blink artifacts were corrected using the algorithm of Gratton et al. (8).

On the basis of the findings of Stockburger et al. (29), we focused on the LPP component of the ERP. The LPP is a late-occurring midline component with an onset after 300 ms, which is larger for items with higher levels of arousal, such as pleasant or unpleasant pictures, relative to neutral stimuli. Separate averages were created for food and flower pictures with an epoch of 200 ms before stimulus presentation and 1000 ms after stimulus presentation. Epochs were baseline corrected using the 200-ms prestimulus window. Electrode sites for analysis were chosen based on the scalp distribution of the current data and the findings of Stockburger et al., which showed the strongest effects of food motivation between 450 and 600 ms. Thus, the LPP was quantified as the mean amplitude between 450 and 600 ms after stimulus presentation averaged across six centroparietal electrode sites 62 (Pz), 67, 71, 72, 76, and 77 (see Larson et al. [18] for figure of EGI sensor net). No latency calculations were made for the LPP because it is a tonic (longer-lasting) and not peaked ERP component.

Energy intake

After completion of the food motivation experimental task, participants were asked to track their energy intake until the following morning during each condition. Each participant was given a food record to log all food and beverages consumed, amount consumed, time of day the food item was consumed, and any other pertinent details about the food or beverage such as the brand name and nutrition labels. To improve the accuracy of food records, participants used food scales (Ohaus, Inc., Parsippany, NJ) to weigh each food item. Each dietary record was analyzed for energy and macronutrient intake using the Food Processor SQL nutrition software (ESHA Research, Inc., Salem, OR).


This was monitored beginning at initial testing (∼8 a.m.) until the following morning (24 h), using a GT1M accelerometer. On the basis of the number of accelerations per unit of time (epochs), sedentary time, intensity of activity, and total PA were determined over each 24-h condition. Accelerometers have previously been shown to have high reliability and were therefore used as an objective measure of PA (34).

Statistical analysis

Analyses were completed using the statistical software PC-SAS (version 9.3; SAS Institute, Inc., Cary, NC). Descriptive data for body weight, BMI, body fat percentage, energy intake, and PA were reported as means and SD. Paired-samples t-tests and separate two-group (normal-weight, obese) × two-condition (nonexercise, exercise) × two-picture (food, flower) repeated-measures ANOVA were used to examine picture valence and arousal ratings. Food motivation was examined using two-group (normal-weight, obese) × two-condition (nonexercise, exercise) × two-picture type (food, flower) mixed-model ANOVA on ERP amplitudes. Tests of simple effects were used to decompose significant main effects and interactions. The relationships between indices of food motivation, BMI, and ERP amplitudes were assessed using zero-order correlations.

To characterize PA data via accelerometry, the following categories from Troiano et al. (30) were used to describe sedentary time, light-intensity activity time, moderate-intensity activity time, and vigorous-intensity activity time: ≤249 counts per minute, 250–2019 counts per minute, 2020–5998 counts per minute, and ≥5999 counts per minute, respectively. Nonwear time was considered any string of zero counts for ≥20 min (26). Mixed models were used to test within BMI group and within condition differences and to test for a BMI group × exercise condition interaction for energy and macronutrient intake and PA level. Control variables included order of condition and nonwear time.

Among the 35 participants who met the study criteria, one participant did not have a full exercise session recorded for their accelerometer data; therefore, those accelerometer data were not able to be included in the analysis. In addition, four participants (two obese) had too few good ERP trials for reliable averages on one of the two testing sessions; these participants were not included in ERP analyses.


Participant characteristics are summarized in Table 1. The normal-weight women had a statistically lower BMI compared with their obese counterparts (F = 12.87, P < 0.001). The mean BMI (kg·m−2) was 22.9 ± 1.4 and 34.0 ± 4.9 for the normal-weight and obese group, respectively. There was no difference in age between groups (F = 0.46, P = 0.649).

For self-reported ratings of valence and arousal for the food and flower pictures, mean valence and arousal ratings did not differ as a function of exercise condition (P > 0.05). There were no significant main effects or interactions for picture valence (P > 0.05). For picture arousal ratings, consistent with previous studies (29), there was a significant main effect of picture type (F = 9.72, P = 0.004); food pictures were rated as more arousing than flower pictures.

Figure 1 displays LPP waveforms by BMI group, exercise condition, and picture type. The waveforms represent the grand average time course of the brain activity at centroparietal electrode sites from 200 ms before picture presentation to 1000 ms after picture presentation. Zero on the x-axis represents the time of picture presentation during the computerized task. BMI groups did not differ as a function of exercise condition or picture type (F = 1.49, P = 0.233). Notably, however, there was a significant exercise session × picture type interaction, with the LPP amplitudes being disproportionately smaller for food pictures than flower pictures during the exercise condition relative to the nonexercise condition (F = 4.25, P = 0.048). There were no additional significant main effects or interactions for LPP.

LPP waveforms for group and picture type as a function of exercise condition. Time is presented in milliseconds. Zero on the x-axis represents the time of picture presentation; y-axis indicates the amplitude of the brain activity in microvolts. The area in the box represents the time period of the mean amplitude used to quantify the LPP. Note the clear difference between food and flower pictures regardless of obesity group in the top panel, but not the bottom panel of the figure. There was a main effect of picture type (F = 9.72, P = 0.004). BMI groups did not differ as a function of exercise condition or picture type (F = 1.49, P = 0.233). There was a significant exercise condition × picture type interaction (F = 4.25, P = 0.048). There were no additional significant main effects or interactions for LPP.

Table 2 reports the dietary intake × BMI group and exercise condition. Energy intake, CHO, fat, protein, and fiber did not vary by BMI group or by exercise condition (P > 0.05). In addition, there was no significant BMI group × exercise condition interaction found for overall energy (F = 0.86, P = 0.361) or macronutrient intake (P > 0.05).

Dietary intake by BMI group and exercise condition.

Table 3 summarizes the 24-h PA levels by BMI group and exercise condition. The exercise condition resulted in significantly more total PA (counts per day) for both the normal-weight and obese women. PA counts were 73% higher for the normal-weight group and 57% higher in the obese group during the exercise condition compared with the nonexercise condition. Furthermore, moderate-intensity activity, vigorous-intensity activity, and MV PA time (min) were significantly greater, and sedentary time was significantly less for both BMI groups during the exercise condition compared with the nonexercise condition (P < 0.05). Finally, there was a significant BMI group × exercise condition interaction for MV PA (F = 4.48, P = 0.043). The obese women performed significantly less MV PA (69 ± 18 min) on the exercise day than the normal-weight women (89 ± 24 min).

PA by BMI group and exercise condition.


This study addressed a notable lack in the scientific literature regarding the effect of exercise on food motivation/food intake, specifically in obese adults (24). This study showed that a 45-min bout of MV PA exercise, not BMI classification, resulted in a disproportionately lower neural response (LPP ERP amplitudes) to pictures of food relative to pictures of flowers immediately after completion of the exercise session (within 1 h). Thus, our hypothesis was partially incorrect regarding the influence of BMI classification on food motivation.

Because this study used a novel measure of food motivation, a direct comparison to other studies is not possible. Notwithstanding, the results of the present study appear consistent with other studies reporting that appetite is suppressed after exercise (22,25,35). Along with physiological control of appetite, Martins et al. (24) reviewed the effect of exercise and appetite regulation and indicated that factors such as, “environmental, psychological, social, and cultural stimuli” can influence food intake and selection. It has also been suggested that dietary restraint and food disinhibition may influence the relationship between exercise and eating behavior (24,32). For example, Visona and George (32) showed that dieters ate more than nondieters did (all of whom had a high level of dietary restraint) over 12 h, after 60 min of treadmill walking. Because the present study did not measure specific hormonal, environmental, social, or cognitive mechanisms, we cannot rule out the possibility that these factors influenced the results.

Importantly, the disproportionately lower food motivation response to exercise compared with nonexercise in the present study was not associated with a difference in total energy or macronutrient intake (regardless of BMI group). Others have similarly reported that acute exercise did not result in a compensatory increase in energy intake (7,24). This finding likely indicates that the effect of a morning-time 45-min bout of MV exercise on food motivation is either transitory or not strong enough to significantly influence energy intake over the course of the entire day. If transitory, it would be interesting to test how long the effect of exercise on food motivation lasted, the commensurate change in hormones that may affect appetite (and potentially food motivation) such as leptin or ghrelin (28,31), and whether or not the energy deficit caused by the exercise or the exercise itself influenced these hormones, as suggested by Hagobian et al. (10).

Hopkins et al. (13) noted that exercise results in a great deal of interindividual variability in factors that may influence appetite and energy intake. The present study showed a modest amount of variability in food motivation. On examination, ∼84% of participants had a stronger neurological response to food than flowers during the nonexercise condition, whereas ∼67% had a stronger response to food than to flowers during the exercise condition. However, consistent with Hopkins et al. (13), we noted considerable variability in the energy intake response among participants in the present study. Approximately 57% of participants (n = 9 obese) showed higher energy intake during the exercise condition and 43% of participants (n = 8 obese) showed higher energy intake during the nonexercise control condition. In other words, roughly half of participants responded to exercise by eating more that day compared with the control day and the other half responded by consuming less on the exercise day. Thus, an additional point made by Hopkins et al. (13) may have application here; namely, that individual differences in biological or behavioral responses to exercise “interact to determine the propensity for weight change.”

A secondary finding of this study is that a planned, supervised bout of exercise significantly influenced total PA and most intensity of PA levels over 24 h. Interestingly, this study showed spontaneous differences in MV PA by BMI classification. Specifically, the obese women obtained less MV PA than the normal-weight women did on the exercise day and, after accounting for the exercise session, less MV PA (−13 min) compared with the nonexercise day. It is possible that this resulted because the exercise session was more physically taxing (higher relative workload) for the obese women and thus less MV PA was obtained for the remainder of the day. Anecdotally, the obese women expressed more discomfort with the exercise than the normal-weight women. In addition, there may have been a cognitive response to the exercise session in which the obese women felt entitled to less activity during the remainder of the day compared with the nonexercise day or their normal-weight counterparts. These possibilities are speculative; nevertheless, what happens for the balance of a day after exercise appears important, particularly in obese women, and warrants consideration when applying PA as a treatment strategy for obesity.

This study had the significant strengths of utilizing objective and sensitive measures, as well as examining the effect of obesity on the outcomes reported. However, there are also limitations. First, attentional processing of food stimuli (i.e., ERP indices of food motivation) was measured immediately after the exercise bout (within 1 h). This indicates an acute response but does not show how long the response persists. Second, energy intake and PA levels were only observed for a 24-h period during each experimental condition. As noted elsewhere (24), studies are needed to examine the effects of exercise on food-seeking behavior over a longer period.

In conclusion, this study demonstrated that a morning-time bout of exercise is associated with reduced neurologically determined food motivation and increased total PA over 24 h compared with a nonexercise condition. This study also highlights the complexity of the PA and food behavior relationship and that individual characteristics may influence some outcomes (e.g., BMI and MV PA) but not others (e.g., BMI and food motivation). Therefore, additional work is needed not only to confirm these findings but also to provide greater understanding regarding the role of individual characteristics and differences in food behavior responses after exercise. Nevertheless, taken as a whole, these findings have positive implications for the role of exercise in weight management.

The authors have no conflict of interest to disclose.

The results of the present study do not constitute endorsement by American College of Sports Medicine.


1. Bailey BW, Tucker LA, Peterson TR, LeCheminant JD. Test–retest reliability of body fat percentage results using dual energy x-ray absorptiometry and the BOD POD. Med Sci Sports Exerc. 2001; 33 (5): 174S.
2. Berthoud HR. Homeostatic and non-homeostatic pathways involved in the control of food intake and energy balance. Obesity (Silver Spring). 2006; 14 (suppl 5): S197–200.
3. De Castro JM. How can eating behavior be regulated in the complex environments of free-living humans? Neurosci Biobehav Rev. 1996; 20 (1): 119–31.
4. Donnelly JE, Blair SN, Jakicic JM, Manore MM, Rankin JW, Smith BK. American College of Sports Medicine Position Stand: appropriate physical activity intervention strategies for weight loss and prevention of weight regain for adults. Med Sci Sports Exerc. 2009; 41 (2): 459–71.
5. Donnelly JE, Smith B, Jacobsen DJ, et al.. The role of exercise for weight loss and maintenance. Best Pract Res Clin Gastroenterol. 2004; 18 (6): 1009–29.
6. Flegal KM, Carroll MD, Kit BK, Ogden CL. Prevalence of obesity and trends in the distribution of body mass index among US adults, 1999–2010. JAMA. 2012; 307 (5): 491–7.
7. George VA, Morganstein A. Effect of moderate intensity exercise on acute energy intake in normal and overweight females. Appetite. 2003; 40 (1): 43–6.
8. Gratton G, Coles MG, Donchin E. A new method for off-line removal of ocular artifact. Electroencephalogr Clin Neurophysiol. 1983; 55 (4): 468–84.
9. Hackbart SJ, LeCheminant JD, Smith JD, Lox CL. The influence of an environmental cue and exercise on food consumption in college students. Int J Exerc Sci. 2009; 2 (2): 94–105.
10. Hagobian TA, Sharoff CG, Braun B. Effects of short-term exercise and energy surplus on hormones related to regulation of energy balance. Metabolism. 2008; 57 (3): 393–8.
11. Hajcak G, MacNamara A, Olvet DM. Event-related potentials, emotion, and emotion regulation: an integrative review. Dev Neuropsychol. 2010; 35 (2): 129–55.
12. Haskell WL, Lee IM, Pate RR, et al.. Physical activity and public health: updated recommendation for adults from the American College of Sports Medicine and the American Heart Association. Med Sci Sports Exerc. 2007; 39 (8): 1423–34.
13. Hopkins M, King NA, Blundell JE. Acute and long-term effects of exercise on appetite control: is there any benefit for weight control? Curr Opin Clin Nutr Metab Care. 2010; 13 (6): 635–40.
14. King NA, Snell L, Smith RD, Blundell JE. Effects of short-term exercise on appetite responses in unrestrained females. Eur J Clin Nutr. 1996; 50 (10): 663–7.
15. Kirk EP, Donnelly JE, Smith BK, et al.. Minimal resistance training improves daily energy expenditure and fat oxidation. Med Sci Sports Exerc. 2009; 41 (5): 1122–9.
16. Kissileff HR, Pi-Sunyer FX, Segal K, Meltzer S, Foelsch PA. Acute effects of exercise on food intake in obese and nonobese women. Am J Clin Nutr. 1990; 52 (2): 240–5.
17. Lang PJ, Bradley MM, Cuthbert BN. International Affective Picture System (IAPS): Technical Manual and Affective Ratings. Gainesville (FL): National Institute of Mental Health Center for the Study of Emotion and Attention; 1997.
18. Larson MJ, Fair JE, Good DA, Baldwin SA. Empathy and error processing. Psychophysiology. 2010; 47 (3): 415–24.
19. LeCheminant JD, Heden T, Smith J, Covington NK. Comparison of energy expenditure, economy, and pedometer counts between normal weight and overweight or obese women during a walking and jogging activity. Eur J Appl Physiol. 2009; 106 (5): 675–82.
20. LeCheminant JD, Tucker LA, Peterson TR, Bailey BW. Differences in body fat percentage measured using dual energy x-ray absorptiometry and the BOD POD in 100 women. Med Sci Sports Exerc. 2001; 33 (5): 174S.
21. Lin PH, Proschan MA, Bray GA, et al.. Estimation of energy requirements in a controlled feeding trial. Am J Clin Nutr. 2003; 77 (3): 639–45.
22. Lluch A, King NA, Blundell JE. Exercise in dietary restrained women: no effect on energy intake but change in hedonic ratings. Eur J Clin Nutr. 1998; 52 (4): 300–7.
23. Maraki M, Tsofliou F, Pitsiladis YP, Malkova D, Mutrie N, Higgins S. Acute effects of a single exercise class on appetite, energy intake and mood. Is there a time of day effect? Appetite. 2005; 45 (3): 272–8.
24. Martins C, Morgan L, Truby H. A review of the effects of exercise on appetite regulation: an obesity perspective. Int J Obes (Lond). 2008; 32 (9): 1337–47.
25. Martins C, Truby H, Morgan LM. Short-term appetite control in response to a 6-week exercise programme in sedentary volunteers. Br J Nutr. 2007; 98 (4): 834–42.
26. Masse LC, Fuemmeler BF, Anderson CB, et al.. Accelerometer data reduction: a comparison of four reduction algorithms on select outcome variables. Med Sci Sports Exerc. 2005; 37 (suppl 11): S544–54.
27. Nijs IM, Muris P, Euser AS, Franken IH. Differences in attention to food and food intake between overweight/obese and normal-weight females under conditions of hunger and satiety. Appetite. 2010; 54 (2): 243–54.
28. Olive JL, Miller GD. Differential effects of maximal- and moderate-intensity runs on plasma leptin in healthy trained subjects. Nutrition. 2001; 17 (5): 365–9.
29. Stockburger J, Schmalzle R, Flaisch T, Bublatzky F, Schupp HT. The impact of hunger on food cue processing: an event-related brain potential study. Neuroimage. 2009; 47 (4): 1819–29.
30. Troiano RP, Berrigan D, Dodd KW, Masse LC, Tilert T, McDowell M. Physical activity in the United States measured by accelerometer. Med Sci Sports Exerc. 2008; 40 (1): 181–8.
31. Vestergaard ET, Dall R, Lange KH, Kjaer M, Christiansen JS, Jorgensen JO. The ghrelin response to exercise before and after growth hormone administration. J Clin Endocrinol Metab. 2007; 92 (1): 297–303.
32. Visona C, George VA. Impact of dieting status and dietary restraint on postexercise energy intake in overweight women. Obes Res. 2002; 10 (1): 1251–8.
33. Wansink B. Environmental factors that increase the food intake and consumption volume of unknowing consumers. Annu Rev Nutr. 2004; 24: 455–79.
34. Welk GJ, Schaben JA, Morrow JR Jr. Reliability of accelerometry-based activity monitors: a generalizability study. Med Sci Sports Exerc. 2004; 36 (9): 1637–45.
35. Westerterp-Plantenga MS, Verwegen CR, Ijedema MJ, Wijckmans NE, Saris WH. Acute effects of exercise or sauna on appetite in obese and nonobese men. Physiol Behav. 1997; 62 (6): 1345–54.


©2012The American College of Sports Medicine