Special Issue: Adolescent Health: Stress, Sleep, and Lifestyle

Journal of International Medical Research 48(3) 1–9 Effect of eating habits on ! The Author(s) 2019 Article reuse guidelines: in adolescents: sagepub.com/journals-permissions DOI: 10.1177/0300060519889738 a study among Chinese journals.sagepub.com/home/imr college students

Qi Xie1,*, Meng-lei Hao2,*, Ling-bing Meng3, Xiao-qin Zuo2, Peng Guo4, Yong Qiu5, Qiang Wang6, Na Zhang7 and Min Lei8

Abstract Objective: Obesity has been linked to cardiovascular disease, diabetes, and osteoarthritis. Obesity and overweight pose a serious threat to human health, with an estimated 190 million overweight and obese people worldwide. Thus, we investigated the influence of certain eating habits on weight among Chinese college students. Methods: We conducted a cross-sectional survey of 536 college students in Shijiazhuang, China. The survey included questions about eating habits. We analyzed the relationship between par- ticipants’ responses and obesity. Results: Sex, residence, speed of eating, number of meals eaten per day, and a diet high in were found to be correlated with obesity. Our results suggest that increasing the number of meals per day, slowing down the pace of eating, and reducing the intake of high-sugar foods have potential benefits for reducing obesity among college students. Conclusions: In the present study, we found that some dietary habits are related to the occur- rence of obesity among college-aged individuals.

1Department of Nutrition, The Fourth Hospital of Hebei Medical University, Shijiazhuang, Hebei Province, China 7Department of Radiation Oncology, The Fourth Hospital 2 Department of Geriatric Medicine, Affiliated Hospital of of Hebei Medical University, Shijiazhuang, Heibei Province, Qinghai University, Xining, Qinghai Province, China China 3 Neurology Department, Beijing Hospital, National 8Department of Nutrition, The Third Hospital of Hebei Center of Gerontology, Dong Dan, Beijing, China Medical University, Shijiazhuang, Hebei Province, China 4Department of Orthopedics, The Fourth Hospital of Hebei Medical University, Shijiazhuang, Heibei Province, China *These authors contributed equally to this work. 5Anesthesiology Department, Beijing Hospital, National Corresponding author: Center of Gerontology, Dong Dan, Beijing, China Min Lei, Department of Nutrition, The Third Hospital of 6Department of Thoracic Surgery, The First Hospital of Hebei Medical University, No. 139 ZiQiang Road, Hebei Medical University, Shijiazhuang, Hebei Province, Shijiazhuang 050051, Hebei Province, China. China Email: [email protected]

Creative Commons Non Commercial CC BY-NC: This article is distributed under the terms of the Creative Commons Attribution-NonCommercial 4.0 License (https://creativecommons.org/licenses/by-nc/4.0/) which permits non-commercial use, reproduction and distribution of the work without further permission provided the original work is attributed as specified on the SAGE and Open Access pages (https://us.sagepub.com/en-us/nam/open-access-at-sage). 2 Journal of International Medical Research

Keywords Diet, eating habits, obesity, college students, cross-sectional survey, China

Date received: 15 August 2019; accepted: 25 October 2019

Introduction college students, an important group of young people, has not previously been Obesity is a common metabolic disease that studied. involves the pathological state of excessive We administered a questionnaire to col- accumulation of , which can lege students in China. We queried their be damaging to human health. In the past dietary habits and preferences via the few decades, the number of obese individu- survey, and then explored the relationship als has increased worldwide and obesity between students’ responses and their BMI. has become an epidemic.1 There are more We aimed to identify eating habits related than 190 million overweight and obese to weight and to then use this information people globally. Compared with 1980, the 2 to improve unhealthy eating habits and number of obese individuals has doubled. some obesity risk factors common among There are 9 million adolescents in the college students. This study provides new United States with obesity. This has led to insights that may be helpful in exploring a substantially increased prevalence of car- the relationship between dietary habits diovascular and cerebrovascular diseases, and obesity in our target age group. Our , and osteoarthritis in ado- results may provide a scientific basis for lescents, seriously affecting their physical 3,4 interventions related to eating habits that and mental health. Factors involved in can help to reduce and prevent obesity adolescent obesity cannot be ignored, such among college students. as the contribution of genes and metabolic diseases.5 However, changes in social cul- ture represent another important factor. Methods Social and economic development have resulted in a large number of sour- Participants ces becoming easily available to teenagers, Our cross-sectional study was conducted at resulting in excessive energy intake.6 In Hebei Medical University in Shijiazhuang, addition, increased social activities among the capital of Hebei Province in China. younger people have led to declining con- From August to September 2017, we sumption and the associated benefits of a recruited students at Hebei Medical healthy family diet.7 Among sociocultural University. The inclusion criteria for partic- lifestyle changes, unhealthy eating patterns ipants were Chinese college students are a main reasons for the increase in obe- between the ages of 18 and 25 years. sity worldwide.8 It has been proven that less Exclusion criteria included chronic diseases frequent consumption of fruits and vegeta- such as diabetes and endocrine diseases bles and excessive intake of high in (e.g., Cushing syndrome or vacuolar sella) sugar content as well as fast foods directly and long-term use of oral medication affect- affect body mass index (BMI).9 However, ing body weight (e.g., estrogen or glucocor- the role of multiple dietary preferences ticoids). All participants gave their informed related to the incidence of obesity among consent for inclusion before participating in Xie et al. 3 the study, and they provided physical exam- no), eating breakfast (yes, no), excessive ination reports from within the previous food consumption resulting in abdominal 1 year. This study was conducted in accor- distension or discomfort (yes, no), eating dance with the Declaration of Helsinki, snacks at night (never, once or twice a and the study protocol was approved by week, three to six times a week, every the Ethics Committee of Hebei Medical day), consuming high- foods (never, University. Participants were provided with once or twice a week, three to six times a a fact sheet and instructed to follow the week, every day), consuming fruits and veg- survey guidelines. A total of 551 question- etables (never, a small amount, a normal naires were collected, with a participation amount), eating dessert (never, once or rate of 91.8% (551/600). After consulting twice a week, three to six times a week, the physical examination reports of partici- every day), drinking sugar-sweetened bever- pants, questionnaires completed by non- ages (never, one or two servings a week, Chinese college students and individuals three to six servings a week, every day), with chronic diseases were excluded. The exercising regularly (never, less than 1 hour remaining questionnaires were considered per week, 1–2 hours per week, daily), and valid and were included in the study. consuming alcohol (never, less than one per week, one or two drinks per Anthropometric measurement week, more than three drinks per week). BMI is useful for estimating body fat and is Statistical analysis calculated by dividing an individual’s weight (kg) by their height in meters All data were entered into a spreadsheet by squared (m2). Height and weight were the researcher and examined for validity. All measured during physical examinations. statistical analyses were performed using Height was measured using a portable IBM SPSS version 23.0 (IBM, Beijing, rangefinder to the nearest 1 cm and weight China). Measurement data are expressed was measured with a calibrated scale to the as mean standard deviation. Pearson’s nearest 0.1 kg. According to World Health chi-squared test was used to investigate the Organization definitions, individuals are relationship between participants’ general classified as underweight with BMI characteristics, dietary habits, and BMI. <18.5 kg/m2, normal weight with BMI The alpha value was set to 0.05 for all tests. between 18.5 and 24.9 kg/m2, and over- weight with BMI 25 kg/m2. Results Self-administered questionnaire Participant characteristics We collected participant information using Among a total 536 participants included in a self-administered questionnaire. Survey this study, 257 male (22.07 3.42, 47.9%) questions included general information: and 279 female (21.10 2.73, 52.1%). class, name, student number, dormitory Participants were age 17 to 22 years, with number, region of residence, and sex; 254 (21.98 3.46, 47.4%) from urban areas eating habits: speed at which meals are and 282 (21.19 2.71, 52.6%) from rural eaten (slow, normal, slightly fast, fast), areas. In terms of college grade, 67 were number of times participants eat per day freshmen (22.41 3.38, 12.5%), 202 were (twice, three times, four times, more than sophomores (21.25 2.74, 37.7%), 177 four times), dining at regular times (yes, were juniors (21.65 3.21, 33.0%), and 90 4 Journal of International Medical Research were seniors (21.46 3.41, 16.8%). Sixty- affected BMI. Second, the number of seven (17.61 0.76, 12.5%) participants meals eaten per day was negatively correlat- were underweight, 396 (21.14 1.69, ed with BMI. Third, the frequency of eating 73.9%) were normal weight, and 73 (27.48 foods high in sugar content, and especially 2.19, 13.6%) were overweight. Value in the speed at which meals are consumed, parentheses are the mean BMI standard were positively correlated with BMI. deviation and rates are the proportion in Sex increases the risk of obesity. Sex hor- each group. Participants from urban areas mones, specifically regulated by related had a higher BMI (p ¼ 0.019) than those genes on the X chromosome, have an from rural areas. important role in maintaining energy bal- ance.11 Participants from urban areas had Relationship between BMI, pace of eating, a higher BMI than those from rural areas. meal frequency, and consumption of Regarding the obvious regional differences in BMI, individuals from urban areas are sugary foods more inclined to smoke and drink alcohol We analyzed the relationship between BMI than those from rural regions. In addition, and participants’ eating habits and found rural residents are more likely to attain that sex, residence, eating speed, daily meal levels of physical activity recommended by frequency, habits of consuming sweets, and the World Health Organization than urban the number of sugar-sweetened beverages residents. Moreover, the intake of sweet or consumed per week were related to BMI. sugary drinks among urban residents is The mean levels for sex (Figure 1a), residence higher than that among residents from (Figure 1b), eating speed (Figure 1c), daily rural areas.12 Thus, individuals from meal frequency (Figure 1d), high-sugar diet urban areas tend to have greater energy (Figure 1e), and drinking sugary beverages intakes than rural residents, which may (Figure 1f) are depicted according to differ- lead to the regional differences in BMI iden- ent BMI categories. Our results indicated tified in our study. that the number of meals eaten per day was The normal feasting–hunger cycle is the negatively correlated with BMI (P ¼ 0.018). primary way that human physiological and Consuming a high-sugar diet was positively metabolic balance is maintained. Satiety correlated with BMI (P ¼ 0.023) but the and appetite are affected by meal frequen- effect on BMI was relatively small. Eating cy, suggesting that appropriate meal fre- speed showed a significant positive correla- quency may help in reducing obesity.13 tion with BMI (P < 0.001). As an individual eats increasingly more, the thermal effect of the consumed food is considerable, resulting in a continuous Discussion 14 increase in the body’s metabolic rate. In the present study, we conducted a ques- These results of past studies are consistent tionnaire survey to assess the weight status with those of our investigation. Meal fre- of college students in China. Based on the quency can also affect gut microbes. results, we explored the influence of certain Intestinal microorganisms indirectly affect eating habits on the likelihood of being leptin secretion by influencing circadian overweight. Shan provided a preliminary rhythms, which take part in the develop- description, and we conducted a further ment of obesity.15,16 Irregular eating fre- analysis of that research.10 We revealed quency also leads to differences in leptin several main findings. First, we found that and ghrelin secretion, reducing energy participants’ sex and region of residence and increasing caloric intake.17 Xie et al. 5

Figure 1 (a) Male and female students, according to body mass index (BMI). (b) Urban and rural residents, according to maximum weight. (c) BMI, according to eating speed. (d) BMI, according to meal frequency. (e) Weekly consumption of sweets among participants, according to BMI. (f) Weekly consumption of sugar- sweetened beverages among participants, according to BMI.

Eating sweets and desserts has a role On the one hand, the binding between insu- in obesity through affecting hormones. lin and its receptors on dopamine neurons is Excessive consumption of sweet foods can reduced, and the feeding regulation func- substantially increase insulin resistance.18 tion of dopamine is weakened.19 On the 6 Journal of International Medical Research other hand, overconsumption of sweets increasing the release of dopamine from may lead to disorders of motivational these neurons.19 behavior and reward-related mechanisms We also found that eating at a normal in individuals who are obese.20 In addition, speed and at a slightly faster speed were dopamine activity and receptor expression more frequent in the normal-weight group. are inhibited, affecting normal dietary reg- Eating at a normal speed and slightly fast ulation.21 Fructose is an indispensable form may lead to a difference in the degree of of sugar that can stimulate the nervous and intervention by endocrine hormones related hedonic responses of the brain and induce to eating, such as insulin and dopamine. leptin resistance.22 However, our study Eating at a normal speed is conducive to findings showed the effect of consuming maintenance of by excessive sweets on obesity to be limited, endocrine hormones and eating at a slightly which may be related to weak roles of insu- faster speed is obviously conducive to pos- lin resistance and intestinal microorganisms itively increasing BMI. Surprisingly, we in obesity. found that a smaller proportion of individ- Meal consumption speed was positively uals with normal weight ate slowly than correlated with obesity, which may be those who ate at a normal speed. There linked to a number of factors. First, eating are several possible reasons for this finding. slowly means longer chewing time, which First, eating at a slow speed may be related increases satiety and reduces excess energy to dissatisfaction with the taste of the 23 intake. Second, eating slowly may reduce food consumed at mealtimes, resulting in intestinal digestive enzyme activity and as a increased snacking between meals.27 result, may reduce food digestion and Second, eating at a normal speed may be absorption. In addition, with extended the set-point for the release of obesity- mealtimes, the brain centers will continu- related endocrine hormones. Slow and fast ously respond to visual information, con- eating speeds may interfere with the release siderably enhancing the food intake of endocrine hormones and affect an indi- memory function and reducing the subse- vidual’s weight. quent intake of food.24 Recent research also suggests that obesity and endocrine Strengths and limitations of this study hormones are closely related. Feeding behavior is often accompanied by olfactory After analyzing the relevant literature in and other physiological signals, which relation to our study findings, we can con- 23 28 increase food intake by stimulating the clude that meal speed, eating frequency, 29 dopaminergic circuit.25 The initial stage of consuming a high-sugar diet, and an indi- the gastrointestinal reflex enables dopamine vidual’s sex30 are all related to obesity. neurons to regulate eating behavior via the However, our study has some limitations. neurohumoral reflex. Spending a long time First, causality cannot be determined eating can lead to loss of control of this using a cross-sectional study design; for regulation mechanism, which destroys the this, prospective and interventional studies normal feeding pattern.26 Insulin secretion are needed. Second, individuals possess a in plasma is pulsed to coincide with food wide variety of eating habits, and we intake. Slowing down the rate of eating is could not include them all in the present beneficial to insulin secretion. In addition to investigation. Finally, our research partici- lowering blood sugar levels, insulin can pass pants were students at Hebei Medical through the blood–brain barrier and bind University in Shijiazhuang between the to insulin receptors on dopamine neurons, ages of 17 and 22 years; therefore, our Xie et al. 7 study population does not represent the Author contributions general population across all age groups Meng-lei Hao and Qi-Xie were the main contrib- in China. Moreover, we examined the rela- utors to the writing and submission of the man- tionship between the speed and frequency uscript. Ling-bing Meng and Min-Lei made of eating meals on obesity, but we did not substantial contributions to the study’s concep- investigate the roles of energy intake per tion. Yong Qiu and Xiao-qin Zuo provided meal or total daily energy intake.31 assistance and suggestions during the submission Finally, the methods used to collect behav- process. Qiang-Wang, Na-Zhang, and Peng Guo ioral data may introduce the risk of bias. obtained and analyzed the data. All authors read and approved the final manuscript.

Conclusions ORCID iD In this study, we explored the relationship Min Lei https://orcid.org/0000-0002-8403- between certain eating habits and the risk of 4771 obesity among Chinese college students. Our results suggested that eating meals References more frequently, slowing down the speed 1. Vecchie A, Dallegri F, Carbone F, et al. of eating, and reducing the intake of Obesity phenotypes and their paradoxical sweets and desserts may have potential ben- association with cardiovascular diseases. efits for reducing obesity among college stu- Eur J Intern Med 2018; 48: 6–17. dents. Our findings provide some useful 2. Saltiel AR and Olefsky JM. Inflammatory information for interventions to prevent mechanisms linking obesity and metabolic the development of obesity and reduce the disease. J Clin Invest 2017; 127: 1–4. incidence of obesity-related diseases, which 3. Ladhani M, Craig JC, Irving M, et al. will have positive effects for individuals and Obesity and the risk of cardiovascular and their families. Our findings provide a scien- all-cause mortality in chronic kidney disease: tific basis for reducing overweight and obe- a systematic review and meta-analysis. sity among college students, as well as new Nephrol Dial Transplant 2017; 32: 439–449. ideas for studying the relationship between 4. Ogden CL, Fryar CD, Hales CM, et al. dietary habits and obesity in other age Differences in obesity prevalence by demographics and urbanization in US chil- groups. dren and adolescents, 2013-2016. JAMA 2018; 319: 2410–2418. DOI: 10.1001/ Acknowledgments jama.2018.5158. We thank Hui-bin Liu (Hebei Medical 5. Ling C and Ronn€ T. Epigenetics in human University) for his assistance with the statistical obesity and type 2 diabetes. Cell Metab analysis and suggestions during the submission 2019; 29: 1028–1044. process. 6. Ebbeling CB, Sinclair KB, Pereira MA, et al. Compensation for energy intake from fast Declaration of conflicting interest food among overweight and lean adoles- The authors declare that there is no conflict of cents. JAMA 2004; 291: 2828–2833. DOI: interest. 10.1001/jama.291.23.2828. 7. Shier V, Nicosia N and Datar A. Neighborhood and home food environment Funding and children’s diet and obesity: evidence This research received no specific grant from any from military personnel’s installation assign- funding agency in the public, commercial, or ment. Soc Sci Med 2016; 158: 122–131. DOI: not-for-profit sectors. 10.1016/j.socscimed.2016.03.043. 8 Journal of International Medical Research

8. Wellington N, Shanmuganathan M, de in patients with type 2 diabetes. Nutrients Souza RJ, et al. Metabolic trajectories fol- 2017; 9: pii: E322. lowing contrasting prudent and western 19. Fordahl SC and Jones SR. High-fat-diet- diets from food provisions: identifying induced deficits in dopamine terminal robust biomarkers of short-term changes in function are reversed by restoring insulin habitual diet. Nutrients 2019; 11: pii: E2407. signaling. ACS Chem Neurosci 2017; 8: DOI: 10.3390/nu11102407. 290–299. 9. Khan MMA, Karim M, Islam AZ, et al. 20. Edwin Thanarajah S, Iglesias S, Prevalence of overweight and obesity Kuzmanovic B, et al. Modulation of mid- among adolescents in Bangladesh: do brain neurocircuitry by intranasal insulin. eating habits and physical activity have a Neuroimage 2019; 194: 120–127. gender differential effect? J Biosoc Sci 21. Jones KT, Woods C, Zhen J, et al. Effects of 2019; 51: 843–856. diet and insulin on dopamine transporter 10. Shan MJ, Zou YF, Guo P, et al. Systematic activity and expression in rat caudate-puta- estimation of BMI: a novel insight into pre- men, nucleus accumbens, and midbrain. dicting overweight/obesity in undergraduates. J Neurochem 2017; 140: 728–740. Medicine (Baltimore) 2019; 98: e15810. DOI: 22. Isganaitis E and Lustig RH. Fast food, cen- 10.1097/md.0000000000015810. tral nervous system insulin resistance, 11. Wang C and Xu Y. Mechanisms for sex dif- and obesity. Arterioscler Thromb Vasc Biol ferences in energy homeostasis. J Mol 2005; 25: 2451–2462. Endocrinol 2019; 62: R129–R143. 23. Shah M, Copeland J, Dart L, et al. Slower 12. Price AJ, Crampin AC, Amberbir A, et al. eating speed lowers energy intake in normal- Prevalence of obesity, hypertension, and dia- weight but not overweight/obese subjects. betes, and cascade of care in sub-Saharan J Acad Nutr Diet 2014; 114: 393–402. Africa: a cross-sectional, population-based 24. Robinson E, Aveyard P, Daley A, et al. study in rural and urban Malawi. Lancet Eating attentively: a systematic review and Diabetes Endocrinol 2018; 6: 208–222. DOI: meta-analysis of the effect of food intake 10.1016/s2213-8587(17)30432-1. memory and awareness on eating. Am J 13. Kerksick CM, Arent S, Schoenfeld BJ, et al. Clin Nutr 2013; 97: 728–742. International society of sports nutrition 25. Thanarajah SE, Backes H, DiFeliceantonio position stand: nutrient timing. J Int Soc AG, et al. Food intake recruits orosensory Sports Nutr 2017; 14: 33. and post-ingestive dopaminergic circuits to 14. Vilela S, Correia D, Severo M, et al. Eating affect eating desire in humans. Cell Metab frequency and weight status in Portuguese 2019; 29: 695–706.e4. children aged 3–9 years: results from the 26. Boekhoudt L, Roelofs TJM, de Jong JW, cross-sectional National Food, Nutrition et al. Does activation of midbrain dopamine and Physical Activity Survey 2015–2016. neurons promote or reduce feeding? Int J Public Health Nutr 2019; 22: 2793–2802. Obes (Lond) 2017; 41: 1131–1140. 15. Voigt RM, Forsyth CB, Green SJ, et al. 27. Koopman KE, Caan MW, Nederveen AJ, Circadian rhythm and the gut microbiome. et al. Hypercaloric diets with increased Int Rev Neurobiol 2016; 131: 193–205. meal frequency, but not meal size, increase 16. Pan W and Kastin AJ. Leptin: a biomarker intrahepatic triglycerides: a randomized for sleep disorders? Sleep Med Rev 2014; 18: controlled trial. Hepatology 2014; 60: 283–290. 545–553. 17. Dashti HS, Scheer FA, Jacques PF, et al. 28. Dallacker M, Hertwig R and Mata J. The Short sleep duration and dietary intake: epi- frequency of family meals and nutritional demiologic evidence, mechanisms, and health health in children: a meta-analysis. Obes implications. Adv Nutr 2015; 6: 648–659. Rev 2018; 19: 638–653. 18. Jung CH and Choi KM. Impact of high- 29. van Dam RM, Rimm EB, Willett WC, et al. carbohydrate diet on metabolic parameters Dietary patterns and risk for type 2 diabetes Xie et al. 9

mellitus in U.S. men. Ann Intern Med 2002; weekend recovery sleep. Curr Biol 2019; 29: 136: 201–209. 957–967.e4. 30. Depner CM, Melanson EL, Eckel RH, et al. 31. Fock KM and Khoo J. Diet and exercise Ad libitum weekend recovery sleep fails to in management of obesity and over- prevent metabolic dysregulation during a weight. J Gastroenterol Hepatol 2013; 28: repeating pattern of insufficient sleep and 59–63.