¹ Department of Nutrition, School of Health, Qazvin University of Medical Sciences, Qazvin, Iran.
² Department of Public Health, School of Health, Qazvin University of Medical Sciences, Qazvin, Iran.
³ Department of Nutrition, School of Paramedicine, Ahvaz Jundishapur University of Medical Sciences, Ahvaz, Iran.
⁴ Department of Statistics and Epidemiology, School of Public Health, Tabriz University of Medical Sciences, Tabriz, Iran.
⁵ Department of Nutrition, School of Public Health, Iran University of Medical Sciences, Tehran, Iran.

 

Obesity, a major public health problem worldwide, is sometimes a true epidemic (Katagiri et al., 2014). The high morbidity related to obesity highlights the importance of recognizing various factors that may increase its high prevalence in the population. Expressed phenotype of obesity is affected by non-genetic determinants including environmental and behavioral factors (Bouchard, 2007, Friedman, 2004, Hill et al., 2003). Physical activity and dietary patterns have obtained much attention as effective factors of obesity over the past decades; however, many other determinants have been recognized to influence the obesity epidemic (Keith et al., 2006). In this regard, sleep deprivation has received much attention (Schoenborn et al., 2004). The literature showed that  sleep quality in short sleepers might be poorer (Hayashino et al., 2007).
Based on cross-sectional and case-control studies, insufficient sleep and poor quality sleep are related to obesity (Gupta et al., 2002, Sekine et al., 2002, Vioque et al., 2000, Von Kries et al., 2002). The possible mechanisms proposed for this association include irregular food intake and lower vegetable consumption (Ohida et al., 2001), lower physical activity due to daytime sleepiness alterations in the balance of energy intake and energy expenditure (Shi et al., 2008), altered levels of orexigenic and anorexigenic hormones (Spiegel et al., 2004, Taheri et al., 2004), and diet quality (Azadbakht and Esmaillzadeh, 2011).
Some investigations demonstrated that acute sleep restriction elevates calorie intake in adults due to elevated consumption of carbohydrates and fats (Bosy-Westphal et al., 2008, Brondel et al., 2010). The results of a study over men sleeping 4 hours for two successive nights reported an elevated appetite particularly for energy dense foods with a high level of carbohydrate (Spiegel et al., 2004).
Although many studies evaluated the relationship of sleep quality and/or duration with weight status and dietary intakes, no study has ever evaluated these associations among Iranian young female students. Thus, the aim of this study was to elucidate the possible negative impacts of poor-quality sleep on the dietary intakes and anthropometric variables among the female students in Qazvin University of Medical Sciences, Iran.
Materials and Methods
Participants and design: This cross-sectional study was carried out among young female students aged 18-30 years. The participants were selected from Qazvin University of Medical Sciences, Qazvin, Iran from February to March 2018. A random sample of 260 healthy female students was selected using a multistage sampling method from schools and departments of Qazvin University of Medical Sciences. Volunteers could participate in this study if they were native single female; aged 18-30 years; had body mass index (BMI) of 18.5 to 35 kg/m²; were under no medication and/or dietary supplements; and were not suffering from mental disability. Shift workers were excluded from the study.
Sleep quality assessment: Sleep quality was investigated using the validated Iranian version of Pittsburg Sleep Quality Index (PSQI) (Naderi et al., 2017). The PSQI is a validated self-report measure of subjective sleep quality and sleep disturbances during the past month period. Briefly, each participant answered nine questions, which generated a score (range 0-3, 0 = no difficulty and
3 = sever difficulty) for seven components of sleep quality, sleep latency, sleep duration, sleep efficacy, sleep disturbances, use of sleeping medication, and day time dysfunction. Sum of these seven scores indicated a global score for sleep quality (range 0-21). A total score of > 5 showed “poor” quality sleep, while participants with total score of ≤5 were categorized as “good” sleepers.
Anthropometric, dietary intake and physical activity measurements: Anthropometric measurements including height, weight, waist circumference, and body fat percentage were conducted for all participants. Height was measured by a non-stretched tape, while the participant was in a standing position and without shoes. Weight was measured using a calibrated digital scale with minimal clothing and without shoes. The participants' BMI was calculated as body weight (kg) divided by height squared (m²). Waist circumference was evaluated at the narrowest region between the lowest rib and the iliac crest over light clothing by a non-stretched tape (Wang et al., 2003). Body fat mass percentage (FM%) was measured by Bioelectrical Impedance Analyzer (Quad scan 4000; Body stat).
Dietary intakes were investigated through a validated 168-item semi-quantitative food frequency questionnaire (FFQ) (Haghighatdoost et al., 2012, Mirmiran et al., 2010), which asked about the portion size and frequency of food and beverage consumption over the past year on a daily, weekly, or monthly basis. The reported frequency for each food item was converted to daily intake. The participants were provided with the potion size images to evaluate the reporting accuracy. Household measures were used to alter all portion sizes of the consumed foods to grams. FFQ was completed by a trained interviewer in 30-36 minutes. Nutrient analysis of diets was conducted by nutritionist 4 software. We evaluated calorie (kcal/day), macronutrients (g/day), and food groups (serving/day) intakes from FFQ. Basically, all food items were classified under five food groups, provided by U.S. Department of Agriculture’s Food Guide Pyramid (bread-grains, vegetables, fruits, meats, and dairy).
Physical activity level was investigated by Persian short form of the International Physical Activity Questionnaire (IPAQ) and reported in Met-Min/week (Craig et al., 2003). A questionnaire was used for socioeconomic status information.
Data analysis: Statistical analysis was performed by SPSS version 21.0 (SPSS Inc, Chicago, Illinois) software. The normal distribution of date
was examined by Kolmogorov-Smirnov test. Furthermore, Mean ± SD was used to describe normal distributed data. Non-normally distributed data was described as Median (interquartile).
The mean values of the dietary intakes, anthropometric indices, and physical activity were compared between good and poor sleepers using independent t-test for normal distributed variables and Mann-Whitney U test for non-normal distributed data. Spearman correlation was used to detect possible bivariate associations. P-value < 0.05 was considered significant.
Ethical considerations: The 260 female students, who met the inclusion criteria, were informed about the aims and protocol of the study, wrote informed consent forms, and
signed them. This study was approved by
the Ethics Committee of Qazvin University
of Medical Sciences, Qazvin, Iran (code: IR.QUMS.REC.1396.413).
Results
Two hundred sixty female students, who were studying in Qazvin University of Medical Sciences, Iran, participated in the current study. The participants' mean age was 21.95 ± 2.11 years. Of 260 female students who participated in the study, 111 (42.7%) were poor sleeper (PSQI > 5) and 149 (57.3%) were good sleeper (PSQI ≤ 5). Table 1 demonstrates anthropometric indices and dietary intakes of the participants. A questionnaire was administered for socioeconomic status information and no significant difference was found between participants in this regard.
Table 2 compares dietary intakes between good and poor sleeper groups. As demonstrated in this table, those in the poor sleep quality group (PSQI > 5) had higher intakes of calorie 2460 versus 2175; P < 0.001) kcal/d and carbohydrate 352 versus 290 g/d; P < 0.001) in comparison with good sleepers (PSQI ≤ 5) and lower protein intake 88 versus 99 g/d; P < 0.01). Moreover, regarding five food groups, poor sleepers had higher intake of bread-grains 9 ± 2 versus 8 ± 2 serving/d; P = 0.02) group, but lower intake of meat-alternatives 5 ± 2 versus 5 ± 1 serving/d; P < 0.001), vegetables 2 ± 1 versus 2 ± 1 serving/d; P < 0.001) and fruits (2 ± 1 versus 3 ± 1; P < 0.001) in comparison to good sleepers (Table 2).  
No significant difference was observed between the two groups regarding the frequency of main meals (breakfast, lunch, and dinner) and between meals per day (P > 0.05) (data not shown).
Participants in the poor quality sleep group (PSQI > 5) had higher weight (65.52 ± 9.64 versus 57.46 ± 7.33 kg; P < 0.001), BMI (25.06 ± 3.30 versus 21.87 ± 2.38 kg/m²; P < 0.001), waist circumference (78.63 ± 7.31 versus 70.70 ± 5.95 cm; P < 0.001), and body FM% (36.08 ± 5.42 versus 29.66 ± 5.19 percent; P < 0.001) in comparison with those in the good sleep quality group (PSQI ≤ 5) (Table 3).
Moreover, a significant difference was observed between good sleepers (1157 Met-Min/week) with poor sleepers (820 Met-Min/week) in relation to physical activity (P < 0.01).
Table 4 and 5 show the correlation between some sleep habits with anthropometric indices, food groups and food intakes in the participants. As these tables demonstrates, total night sleep had a positive and significant correlation with intake of vegetables per day (r = 0.01, P < 0.01), and total sleep time and intake of vegetables per day (r = 0.12, P = 0.04). Furthermore, the sleep onset latency had a significant association with weight (r = 0.26, P < 0.001), body fat mass percentage (r = 0.30, P < 0.001), waist circumference (r = 0.28, P < 0.001), calorie intake (r = 0.15, P = 0.01), protein intake (r = 0.17, P < 0.01), carbohydrate intake (r = 0.16, P < 0.01), bread-grain group intake (r = 0.13, P = 0.03), and the number of meat serving intake per day (r = 0.25, P < 0.001). In addition, sleep onset latency had a negative
and significant correlation with intake of vegetable (r = -0.35, P < 0.001) and fruits
(r = -0.34, P < 0.001) per day. No significant correlation was observed between other variables (P ≥ 0.05).

 

^a: Values are shown as median (Inter Quartiles)
 

 

 

A: Calorie (kcal/d), B: Protein (g/d), C: Carbohydrate (g/d), D: Fat (g/d), E: Bread-grain (serving/d),
F: Vegetables (serving/d), G: Fruits (serving/d), H: Meats (serving/d), I: Dairy (serving/d)