Daytime Sleepiness , Circadian Preference , Caffeine Consumption and Use of Other Stimulants among Thai College Students

We conducted this study to evaluate the prevalence of daytime sleepiness and evening chronotype, and to assess the extent to which both are associated with the use of caffeinated stimulants among 3,000 Thai college students. Demographic and behavioral characteristics were collected using a self-administered questionnaire. The Epworth Sleepiness Scale and the Horne and Ostberg Morningness-Eveningness Questionnaire were used to evaluate prevalence of daytime sleepiness and circadian preference. Multivariable logistic regression models were used to evaluate the association between sleep disorders and consumption of caffeinated beverages. Overall, the prevalence of daytime sleepiness was 27.9 % (95% CI: 26.2–29.5%) while the prevalence of evening chronotype was 13% (95% CI: 11.8–14.2%). Students who use energy drinks were more likely to be evening types. For instance, the use of M100/M150 energy drinks was associated with a more than 3-fold increased odds of evening chronotype (OR 3.50; 95% CI 1.90–6.44), while Red Bull users were more than twice as likely to have evening chronotype (OR 2.39; 95% CI 1.02– 5.58). Additionally, those who consumed any energy drinks were more likely to be daytime sleepers. For example, Red Bull (OR 1.72; 95% CI 1.08–2.75) or M100/M150 (OR 1.52; 95% CI 1.10–2.11) consumption was associated with increased odds of daytime sleepiness. Our findings emphasize the importance of implementing educational and prevention programs targeted toward improving sleep hygiene and reducing the consumption of energy drinks among young adults Corresponding author. Dr. Bizu Gelaye, Department of Epidemiology, Harvard School of Public Health, 677 Huntington Ave, K501, Boston, MA 02115 USA, Telephone: 617-432-6477, Facsimile: 617-566-7805, bgelaye@hsph.harvard.edu. Conflict of Interest The authors have no competing interests to declare. NIH Public Access Author Manuscript J Public Health Epidemiol. Author manuscript; available in PMC 2014 October 27. Published in final edited form as: J Public Health Epidemiol. 2014 June ; 8(6): 202–210. doi:10.5897/JPHE2014.0620. N IH -P A A uhor M anscript N IH -P A A uhor M anscript N IH -P A A uhor M anscript Introduction Sleep is essential for maintaining good health in humans1. However, adolescents experience many changes in their sleep patterns in part due to physiological delayed sleep phase syndrome and disruption of the homeostatic sleep-wake cycle2, 3. The challenges associated with the transition to college, including reduced parental supervision extracurricular activities, academic workloads, and social commitments2–4, adversely affect sleep duration and sleep patterns of college students2, 5. Irregular sleep patterns can lead to daytime sleepiness, which is a result of sleep loss and deprivation6, 7. There is a substantial body of evidence that shows daytime sleepiness is associated with reduced emotional intelligence, impaired constructive thinking skills, poor academic performance, job loss, headaches, and obesity7–9. Additionally, circadian rhythms and preferences can also be affected by social, biological, and environmental factors10–12. There are two major types of circadian preferences: morningness and eveningness13, 14. Morning types tend to wake up relatively early in the morning and work best during this time13. On the other hand, evening types wake up later in the day and prefer to work during the evening and night13. An accumulating body of literature has confirmed that evening types experience higher sleep apnea, increased stress hormones, less healthier lifestyles, and lower academic performances, although much of the research has focused on Western populations10–12, 15. Caffeinated beverages and energy drinks have been implicated as important risk factors for increased daytime sleepiness and evening chronotype among college students16, 17. Globally, energy drinks have gained popularity among adolescents and young adults to counteract tiredness and meet academic, physical, and cognitive demands4, 18–20. Some adverse effects of caffeine intake include energy loss, headaches, cardiac problems, and even sudden death19, 21, 22. Despite the rising trends and aggressive marketing strategies aimed toward college students, there has been little research done on college students’ intake of energy drinks and their impact on sleep disorders4, 23. In particular, Thailand has one of highest rates of energy drink consumption per capita, but very few studies have focused on sleep-related health problems in Southeast Asian populations24–28. Given documented relationships between energy drinks and sleep disorders, we hypothesized that students who use more stimulants were more likely than non-users to experience daytime sleepiness16, 19, 29. We also hypothesized that students who use stimulants are more likely than non-users to be evening types17. From this study, we expect to provide evidence that will guide the development of health and wellness programs for young adults in Thailand. Methods and Materials Study Setting and Sample This cross-sectional study was conducted between December 2010 and February 2011 at seven private and government colleges in Thailand30. Details of the study setting, sampling, and data collection procedures have been described previously30. A total of 3,000 Tran et al. Page 2 J Public Health Epidemiol. Author manuscript; available in PMC 2014 October 27. N IH -P A A uhor M anscript N IH -P A A uhor M anscript N IH -P A A uhor M anscript undergraduate students participated in the study and completed a self-administered survey. All study questionnaires were anonymous, and no personal identifiers were collected. Study procedures were approved by the institutional review boards of the Faculty of Medicine Chulalongkorn University, Walailak University, and the University of Washington, USA. The Harvard School of Public Health Office of Human Research Administration, USA, granted approval to use the anonymous data set for analysis. Data Collection and Covariates Demographics—In this study, a self-administered questionnaire was used to collect information on demographics and lifestyle characteristics. These include age, sex, education level, smoking status, physical activity, caffeine use, and alcohol consumption. The student’s height, weight, waist, and hip circumference were measured by research nurses after the questionnaire was administered. Use of Stimulant Beverages and Other Caffeinated Drinks—Participants were first asked if they consumed any stimulant or energy drink during the past week. Participants answering “Yes” were further asked to identify the specific types of energy drinks and/or stimulant drinks. These are beverages used to provide an extra boost in energy, promote wakefulness, and provide cognitive and mood enhancement16. In this study, we will be using the terms energy drinks and stimulant beverages interchangeably. To provide a range of popular energy drinks in the immediate geographic regions where the survey was administered, we included examples of energy drinks that were common on the campuses and in surrounding social establishments. These included global and local brands such as Red Bull, M100, M150, CarabaoDaeng, Lipovitan-D or Lipo, Wrangyer, and Sharks. For the purpose of this analysis, we combined energy drinks that were less commonly used (i.e., CarabaoDaeng, Lipovitan-D or Lipo, Wrangyer, and Shark) and classified them as “other energy drinks”. Consumption of other caffeine-containing beverages included coffee, black tea, and stimulant beverages such as Coke and Pepsi or sugar-free Coke and Pepsi. Each caffeinated beverage was dichotomized (yes vs. no). Other Covariates—Alcohol consumption was defined as low (<1 alcoholic beverage a week), moderate (1–19 alcoholic beverages a week), and high to excessive consumption (>19 alcoholic beverages a week)31. Other covariates considered were: age (years), sex, cigarette smoking history (never, former, current), and participation in moderate or vigorous physical activity (no vs. yes); BMI was calculated as weight (in kilograms) divided by height (in meters) squared. BMI thresholds were set according to previously defined WHO cut points (underweight, <18.5 kg/m2; normal, 18.5–24.9 kg/m2; overweight, 25.0– 29.9 kg/m2; and obese, 30 kg/m2)32. Epworth Sleepiness Scale (ESS)—The Epworth Sleepiness Scale (ESS) was used to evaluate general level of daytime sleepiness33 and the capability to stay alert and awake during crucial moments of the day7. The ESS is a brief instrument that has been widely used globally among different study populations. It has 8 items capturing an individual’s propensity to fall asleep during commonly encountered situations, each measured on a scale from 0 to 3. The scores for the eight questions are added together to obtain a single total Tran et al. Page 3 J Public Health Epidemiol. Author manuscript; available in PMC 2014 October 27. N IH -P A A uhor M anscript N IH -P A A uhor M anscript N IH -P A A uhor M anscript score that ranges from 0 to 24. In adults, an ESS score ≥10 is taken to indicate increased daytime sleepiness33. In this study we used the ESS ≥10 cut point to define daytime sleepiness. The Horne and Ostberg Morningness-Eveningness Questionnaire (MEQ)— Circadian preference was assessed using the Horne and Ostberg Morningness-Eveningness Questionnaire14. Used globally, the MEQ14 is a 19-item questionnaire that identifies morningness-eveningess preference13, 15. Circadian rhythm is an individual’s endogenous sleep-wake state during a 24-hour period, while circadian preference refers to a person’s inclination to sleep and engage in activities during that same period13, 14, 34. The scores range from 16 to 86 and participants can be classified in five categories: definite and moderate evening (E)-type, neutral type, and moderate and definite morning (M)-type. Higher values on MEQ indicate stronger morningness preference. In this study we u


INTRODUCTION
Sleep is essential for maintaining good health in humans (Luyster et al., 2012). However, adolescents experience many changes in their sleep patterns in part due to physiological delayed sleep phase and disruption of the homeostatic sleep-wake cycle (Millman, 2005;Oginska and Pokorski, 2006). The challenges associated with the transition to college, including reduced parental supervision, extracurricular activities, academic workloads and social commitments (Malinauskas et al., 2007;Millman, 2005;Oginska et al., 2006) adversely affect sleep duration and sleep patterns of college students (Lund et al., 2010;Millman, 2005). Irregular sleep pattern scan lead to daytime sleepiness, which is a result of sleep loss and deprivation (AASM, 2001;Kushida, 2006). There is a substantial body of evidence that shows daytime sleepiness is associated with reduced emotional intelligence, impaired constructive thinking skills, poor academic performance, job loss, headaches and obesity (AASM, 2001;Killgore et al., 2008;Slater et al., 2013). Additionally, circadian rhythms and preferences can also be affected by social, biological and environmental factors (Besoluk et al., 2011;Kanerva et al., 2012;Lucassen et al., 2013). There are two major types of circadian preferences: morningness and eveningness (Horne and Ostberg, 1976;Schneider et al., 2011). Morning types tend to wake up relatively early in the morning and work best during this time (Schneider et al., 2011). On the other hand, evening types wake up later in the day and prefer to work during the evening and night (Schneider et al., 2011). An accumulating body of literature has confirmed that evening types experience higher sleep apnea, increased stress hormones, less healthier lifestyles and lower academic performances, although much of the research has focused on Western populations (Besoluk et al., 2011;Kanerva et al., 2012;Lucassen et al., 2013;Merikanto et al., 2013).
Caffeinated beverages and energy drinks have been implicated as important risk factors for increased daytime sleepiness and evening chronotype among college students (Ishak et al., 2012;Taillard et al., 1999). Globally, energy drinks have gained popularity among adolescents and young adults to counteract tiredness and meet academic, physical and cognitive demands (Buxton and Hagan, 2012;Malinauskas et al., 2007;Pomeranz et al., 2013;Roehrs and Roth, 2008). Some adverse effects of caffeine intake include energy loss, headaches, cardiac problems and even sudden death (Cannon et al., 2001;Roehrs and Roth, 2008;Seifert et al., 2011), a latest report by global drinks consultancy Zenith International. Thailand, home to the energy drinks concept, still leads in consumption per person.
Despite the rising trends and aggressive marketing strategies aimed toward college students, there has been little research done on college students' intake of energy drinks and their impact on sleep disorders (Arria et al., 2010;Malinauskas et al., 2007). In particular, Thailand leads in consumption of energy drink consumption per capita globally, but very few studies have focused on sleep-related health problems in Southeast Asian populations (Cheng et al., 2012;Doi et al., 2003;Mak et al., 2010;Tsai and Li, 2004;Zenith International, 2007). Given docu-mented relationships between energy drinks and sleep disorders, we hypothesized that students who use more stimulants were more likely than non-users to experience daytime sleepiness (Ishak et al., 2012;James et al., 2011;Roehrs and Roth 2008). We also hypothesized that students who use stimulants are more likely than non-users to be evening types (Taillard et al., 1999). From this study, we expect to provide evidence that will guide the development of health and wellness programs for young adults in Thailand.

Study setting and sample
This cross-sectional study was conducted between December, 2010 and February, 2011 at seven private and government colleges in Thailand (Lohsoonthorn et al., 2012). Details of the study setting, sampling and data collection procedures have been described previously (Lohsoonthorn et al., 2012). Briefly, flyers were posted in each campus to recruit and invite participants to the study. Students who expressed an interest in participating were asked to meet in a large classroom or an auditorium where they were informed about the purpose of the study. Students consenting to participate were asked to complete a self-administered individual survey. Vision impaired students and those who could not read the consent and questionnaire forms were not eligible to participate. Those enrolled in correspondence, extension or night school programs were not included as well since their experience might be different from regular time students. Approximately 90% students invited to participate in the survey elected to do so. A total of 3,000 undergraduate students participated in the study and completed a self-administered survey. All study questionnaires were anonymous, and no personal identifiers were collected. Study procedures were approved by the institutional review boards of the Faculty of Medicine Chulalongkorn University, Walailak University and the University of Washington, USA. The Harvard School of Public Health Office of Human Research Administration, USA, granted approval to use the anonymous data set for analysis.

Demographics
In this study, a self-administered questionnaire was used to collect information on demographics and lifestyle characteristics. These include age, sex, education level, smoking status, physical activity, caffeine use, and alcohol consumption. The student's height, weight, waist and hip circumference were measured by research nurses after the questionnaire was administered.

Use of stimulant beverages and other caffeinated drinks
Participants were first asked if they consumed any stimulant or energy drink during the past week. Participants answering "Yes" were further asked to identify the specific types of energy drinks and/or stimulant drinks. These are beverages used to provide an extra boost in energy, promote wakefulness and provide cognitive and mood enhancement (Ishak et al., 2012). In this study, we will be using the terms energy drinks and stimulant beverages interchangeably. To provide a range of popular energy drinks in the immediate geographic regions where the survey was administered, we included examples of energy drinks that were common on the campuses and in surrounding social establishments. These included global and local brands such as Red Bull, M100, M150, CarabaoDaeng, Lipovitan-D or Lipo, Wrangyer and Sharks. For the purpose of this analysis, we combined energy drinks that were less commonly used (that is, CarabaoDaeng, Lipovitan-D or Lipo, Wrangyer and Shark) and classified them as "other energy drinks". Consumption of other caffeine-containing beverages included coffee, black tea and stimulant beverages such as Coke and Pepsi or sugar-free Coke and Pepsi (contain about 35 mg of caffeine). Each caffeinated beverage was dichotomized (yes vs. no).

Other covariates
Alcohol consumption was defined as low (<1 alcoholic beverage a week), moderate (1 to 19 alcoholic beverages a week) and high to excessive consumption (>19 alcoholic beverages a week) (World Health Organization (WHO), 2004). Other covariates considered were: age (years), sex, cigarette smoking history (never, former, current) and participation in moderate or vigorous physical activity (no vs. yes); Body mass index (BMI) was calculated as weight (in kilograms) divided by height (in meters) squared. BMI thresholds were set according to previously defined WHO cutoff points (underweight, <18.5 kg/m 2 ; normal, 18.5 to 24.9 kg/m 2 ; overweight, 25.0 to 29.9 kg/m 2 ; and obese, 30 kg/m 2 ) (WHO, 1995).

Epworth sleepiness scale (ESS)
The Epworth sleepiness scale (ESS) was used to evaluate general level of daytime sleepiness (Johns, 1991) and the capability to stay alert and awake during crucial moments of the day (AASM, 2001). The ESS is a brief instrument that has been widely used globally among different study populations. It has 8 items capturing an individual's propensity to fall asleep during commonly encountered situations, each measured on a scale from 0 to 3. The scores for the eight questions are added together to obtain a single total score that ranges from 0 to 24. In adults, an ESS score ≥ 10 is taken to indicate increased daytime sleepiness (Johns, 1991). In this study we used the ESS ≥ 10 cutoff point to define daytime sleepiness.

The horne and Ostbergmorningness-eveningness questionnaire (MEQ)
Circadian preference was assessed using the Horne and Ostbergmorningness-eveningness questionnaire (Horne et al., 1976). Used globally, the MEQ (Horne et al., 1976) is a 19-item questionnaire that identifies morningness-eveningess preference (Merikanto et al., 2013;Schneider et al., 2011). Circadian rhythm is an individual's endogenous sleep-wake state during a 24-h period, while circadian preference refers to a person's inclination to sleep and engage in activities during that same period (Barion and Zee, 2007;Horne et al., 1976;Schneider et al., 2011

Statistical analysis
We first examined the frequency distributions of socio-demographic and behavioral characteristics of the study participants. Characteristics were summarized using means (± standard deviation) for continuous variables with normal distribution and counts and percentages for categorical variables. Mean (± standard deviation) of MEQ and ESS scores were calculated across sociodemographic and behavioral characteristics and the associations were tested using a one-way ANOVA for multi-level characteristics and a two-sample t-test for two-level characteristics. Multivariable linear regression models were also fitted to evaluate the associations. We also calculated the distribution of morningnesseveningness chronotype across demographic and behavioral characteristics and Chi-square tests were used to determine bivariate differences. Additionally, we calculated the distribution of daytime sleepiness and evening chronotype across energy drink consumption status. Multivariable logistic regression models were used to calculate odds ratios (ORs) and 95% confidence intervals (95% CIs) for the associations. All analysis were performed using SPSS Statistical Software for Windows (IBM SPSS, version 20, Chicago, IL, USA). All reported p-values are two-sided and deemed statistically significant at a 0.05 level.

RESULTS
Of the 3,000 college students who completed the survey and met participant guidelines, 66.9% of the students were females and the average reported age was 20.3 ± 1.3 years (Table 1). About two-thirds of students reported never drinking, while 32.1% admitted drinking 1 to 19 drinks per month; very few (1.7%) used alcohol excessively. A large majority of students have normal BMI (68.9%), few are overweight (10%) and 16.5% are underweight. Over three-fourths of study participants reported engaging in some type of physical activity. Table  2 shows the prevalence estimates of chronotype levels according to the Horne and Ostberg criteria. Thirteen percent of students were classified as evening types (12% in females and 15.1% in males) while 18.7% were classified as morning types (19.7% in females and 16.5% in males).
The prevalence of daytime sleepiness across age groups and sex is displayed in Figure 1. Daytime sleepiness (ESS ≥ 10) was present in 27.9% of the students (95% CI: 26.2 to 29.5%). Overall males in the 18 and the 22-and-over age groups appear to have higher prevalence of daytime sleepiness compared to females; 27.6% of the 18 year-old males experience daytime sleepiness compared to 21.2% of females; 35% of males 22-and-over experience daytime sleepiness compared to 27.4% of females. For the other age groups, the prevalence of daytime sleepiness is higher among females. Looking at the distribution of ESS total score across sex, the median ESS total score of females are higher than males.
As shown in Table 3, females (vs. males) and smokers (current and former vs. non-smokers) had a significantly lower MEQ score (p value < 0.001). There is also a significant association between MEQ score and alcohol consumption (p value < 0.001); we noted a trend lower MEQ scores with higher levels of alcohol consumption. Age, obesity status and participation in physical activity were not found to be statistically significantly associated with MEQ scores. The multivariable linear regression model with all the demographic and lifestyle characteristics gave similar results, except that sex was no longer found to be significant (Table 1). As shown in Table 4, females (vs. males), alcohol consumers (vs. non-consumers), physically active respondents (vs. non-physically active respondents) and non-obese individuals (vs. obese individuals) had a significantly higher ESS score (p value < 0.037). Age and smoking status was not found to be significantly associated with ESS score. The results from the multivariable linear regression model were not appreciably different from the univariate analyses (Table 2).  Table 5 summarizes the logistic regression results. Consumers of any stimulant beverage had 2.68-folds higher odds of being evening chronotypes compared to those who abstained from consuming stimulant beverages (OR 2.68; 95% CI 2.01 to 3.58), after adjusting for age, sex, smoking, BMI and physical activity. When considering specific types of beverages, the odds of being evening chronotype were between 1.95-and 3.5fold higher among users compared to non-users. Compared to those who consumed less than one stimulant beverages per week, those who consumed two per week have 2.65-folds higher odds of being evening chronotypes (OR 2.65; 95% CI 1.81 to 3.90) while those who consumed three or more have 3.65-folds higher odds of being evening chronotypes (OR 3.65; 95% CI 2.58-5.16). However, consuming one stimulant per week was not statistically significantly associated with the odds of being evening chronotypes (OR 1.41; 95% CI .87 to 2.29). Consumers of stimulant beverages had a 22% higher odds When considering specific types of beverages, the odds of experiencing daytime sleepiness were between 1.21and 1.72-fold higher among users compared to nonusers; exceptions included coffee drinkers or consumers of other types of energy drinks, for whom no statistically significant association were found. Those who consumed three or more stimulants per week had 37% higher odds of experiencing daytime sleepiness compared to those who do not use stimulants (OR 1.37; 95% CI 1.13 to 1.67).

DISCUSSION
To the best of our knowledge, this is the first study to examine the prevalence of daytime sleepiness and evening  evening chronotype in relation to caffeine use in a Southeast Asian population. Approximately 28% experienced daytime sleepiness while 13% of our student cohort reported to be evening types. Overall, students with daytime sleepiness were more likely to be cigarette smokers, alcohol drinkers, obese, physically active and stimulant beverage consumers. The odds of being evening types were elevated among cigarette smokers, alcohol consumers and caffeinated stimulant users. Our results are in accordance with previous reports indicating prevalence of daytime sleepiness and eveningness chronotype and extend this literature to assess their associations with consumption of stimulant beverages in Thai young adults (Adan and Natale, 2002;Besoluk et al., 2011;Chung and Cheung, 2008;Hsu et al., 2012;Lund et al., 2010;Schneider et al., 2011;Wu et al., 2012;Yang et al., 2003). For instance, our result showing that 13% of Thai college students were evening types is consistent with reports from Chung and Cheung (2008) who reported a 13.1% prevalence of evening chronotype among Chinese students. Additionally, our findings showing a 27.9% daytime sleepiness are similar to those of Wu et al. (2012) who reported a 22.2% prevalence of daytime sleepiness. Other investigators, however, have reported higher prevalence estimates (Pirrallo et al., 2012). On balance, the results of our study and those of others emphasize the growing problem of daytime sleepiness and eveningness chronotype among college students.
In our study of Thai college students, consuming any stimulant beverage was strongly associated with evening chronotype and daytime sleepiness. Prior studies have found similar results in other university settings (Giannotti et al., 2002;James et al., 2011;Roehrs and Roth 2008;Snel and Lorist, 2011;Taylor et al., 2011). College marks the beginning of new and stressful changes: increased academic workload, busier social lives and later bed times (Lund et al., 2010;Malinauskas et al., 2007). Several investigators determined that academic stress interferes with sleeping schedule, as students may stay up late into the night to study (Chung et al., 2008;Lund et al., 2010). Given the high prevalence of poor sleep quality, short sleep duration and high rates of energy drink consumption among Thai college students, (Steptoe et al., 2006;Zenith International, 2007) it may be fitting to note that these undergraduates may sacrifice sleep for academic purposes and social commitments and drink caffeinated stimulants to remain alert (Ishak et al., 2012;Malinauskas et al., 2007;Roehrs and Roth 2008). As they stay up later to finish their work, they can become sleep deprived due to the morning-oriented schedules of college and thus experience daytime sleepiness (Kushida, 2006;Wittmann et al., 2010).
Several investigators have noted that evening types and daytime sleepers are more likely to have poor health (AASM, 2001;Kanerva et al., 2012;Killgore et al., 2008;Nakade et al., 2009;Schneider et al., 2011;Slater et al., 2013;Stroe et al., 2010;Wittmann et al., 2010). Our findings suggest that caffeine consumers were more likely to experience daytime sleepiness and be evening types. As a result, caffeine could possibly be linked to other poor health behaviors. Future research must evaluate the extent to which caffeinated beverages and energy drinks are associated with poor health traits.
The association between caffeinated drinks and circadian disruption and daytime sleepiness can be explained by the biological mechanism of melatonin suppression and adenosine blockage. During waking hours, light is known to suppress melatonin production (Barion et al., 2007;Gooley et al., 2011). Given the demanding college workload and social commitments, students may stay up later into the night to study, which requires the use of lighting and encourages the consumption of caffeinated beverages to increase alertness. This increased light exposure leads to melatonin suppression, in which light exposure at night shifts one's chronotype towards eveningness (Barion et al., 2007;Shanahan et al., 1999). Another mechanism that can explain sleep disruption is caffeine's role in adenosine blockage. Within the basal forebrain, adenosine, an endogenous biochemical compound, regulates sleep by inhibiting the cholinergic neurons that create arousal (Basheer et al., 2004;Ribeiro and Sebastiao, 2010;Roehrs and Roth 2008). As caffeine acts as an adenosine blocker, it negates the effects of adenosine to induce sleep (Roehrs and Roth 2008). Consequently, consuming caffeine at night prolongs wakefulness and decreases sleep duration, resulting in daytime sleepiness (James et al., 2011;Kushida, 2006;Roehrs and Roth 2008).
The results from our study should be interpreted in the context of some limitations. First, our study could be subjected to volunteer bias, because the data were collected from willing participants instead of a random sample. Second, the temporal relationship between lifestyle characteristics and sleep disorders cannot be delineated due to the cross-sectional study design. It is possible that daytime sleepiness and enveningess chronotopye might have led to increased consumption of energy drinks and other caffeinated beverages. Prospective studies are required to confirm and expand upon our observations. Third, there may be lifestyle and dietary traits that are heterogeneous within daytime sleepers and evening types, which could affect the strength of the associations between sleep disorders and lifestyle traits. Fourth, our study population was exclusively comprised of participants who were attending college, thus conclusions cannot be other broader populations. Lastly, we did not have information concerning frequency, timing and dose of energy drink consumption in the present study. As a result, it is possible that the binary grouping of energy drink consumption may have attenuated the magnitude of the association toward the null.

Conclusion
In summary, our findings underline the growing problem of energy drinks and their impact on both circadian preference and daytime sleepiness among Thai college students. Despite academics being the frequent reason given for energy drink consumption, evening types or daytime sleepers do not exhibit higher academic performance compared to their counterparts (Besoluk et al., 2011;Chung et al., 2008;Giannotti et al., 2002;James et al., 2011;Taylor et al., 2011).

RECOMMENDATION
There is a large body of evidence that daytime sleepers and evening types are strongly associated with numerous health and social problems (AASM, 2001;Kanerva et al., 2012;Killgore et al., 2008;Nakade et al., 2009;Schneider et al., 2011;Slater et al., 2013). From a public health promotion and disease prevention standpoint, these findings suggest an obvious need for effective educational and prevention programs targeted toward improving sleep hygiene and reducing consumption of energy drinks among young adults. Future research that evaluates the impact of caffeinated beverages on various sleep disorders and those that assess the effectiveness of health and wellness programs among young adults, are needed.