Relations of Dietary patterns and sleep disorders in shift workers

Background & aims Few studies have been done on the association of dietary intakes with SDs among shift workers, in particular those exposed to air pollution. We aimed to investigate the association of major dietary patterns with risk of SDs in shift workers of a cement factory in Iran. Methods This cross-sectional study was carried out on 257 male shift workers in a cement factory in Iran. Usual dietary intake of participants was assessed using a validated self-administrated 147-item dish-based FFQ. To examine sleep disturbances of participants, we used the Epworth Sleepiness Scale (ESS). Results Three major dietary patterns including Classic Iranian (CI), Modern Iranian (MI), and High Protein Low Fruit and vegetables (HPLF) were identied. The CI pattern was mainly consisted of high-protein animal foods, fruit and vegetables, while, the MI pattern had high amounts of rened grains, solid oils, and caffeine drinks. No signicant association was found between adherence to the CI pattern and risk of SDs (OR: 2.33; 95% CI: 0.69, 7.81) and also for the MI dietary pattern (OR: 0.62; 95% CI: 0.20, 1.93). However, we found a signicant direct association between the HPLF dietary pattern and risk of SDs (OR: 3.92; 95% CI: 1.03, 14.86). This study showed a signicant direct association between a dietary pattern high in animal proteins and solid oils and low in fruit and vegetables with risk of sleep disorders among shift workers in the cement factory. However, no such association was found for the two other dietary patterns.


Introduction
Sleep Disorders (SDs) are common among shift workers (1). Shift working disturbs normal sleep in these subjects (2). Disturbances in circadian rhythm might cause metabolic disorders such as hypertension, dyslipidemia, hyperglycemia, and abdominal obesity (3). Cement factory shift workers are frequently suffered from SDs, which are intensi ed by air pollution in such environment (4). SDs cause various health problems in these subjects and will reduce their ability to work (5).
Unhealthy dietary intakes are among the most important risk factors of SDs (6,7). High intake of fruit and vegetables has been associated to the lowered risk of SDs (8,9). In contrast, a signi cant direct association has been found between high intake of red meat and processed foods, as well as fried foods with risk of SDs (10). In addition, some studies have shown that SDs were less common among those with frequent caffeine intake (11). However, some other studies showed signi cant increase in SDs following high intake of caffeine (12). Shift workers are susceptible to SDs because of their working programs (1). In addition, air pollution is one of the factors causing SDs in cement factory workers (4).
Shift working also alters the dietary intake of workers, which might result in several mental disorders (13).
Adherence to a healthy dietary pattern, characterized by high intakes of vegetables, mushrooms, potatoes, seaweeds, soy products, and eggs, once or more a week in a study among Japanese workers, was associated with a decreased prevalence of di culty in sleep (14). In another study in Japanese workers, low intake of vegetables and sh, high intake of sweets and noodles and unhealthy eating habits were independently associated with poor sleep quality. Poor sleep quality was also associated with high carbohydrate intake in that study (15).
Few studies have been done on the association of dietary intakes with SDs among workers, in particular shift workers. In addition, this association might be in uenced by the environment of the workplace.
Moreover, studies on the association of dietary pattern with sleep disorders are commonly done among western populations, we all know that dietary patterns are greatly different between populations. Therefore, we aimed to investigate the association of major dietary patterns with risk of SDs in shift workers of a cement factory in Iran.

Study design and participants
This cross-sectional study was carried out on 257 male workers in one of the biggest cement factories in Iran. Participants had at least a 6-month experience and were 20-60 years old. All workers examined were shift workers with shift periods of three weeks. The rst two weeks they work on alternate days from 3 pm to 7 am, and the last week their work schedule is daily from 7 am to 3 pm. Shift workers who had a body mass index of less than 40 kg /m 2 , without kidney, liver, cardiovascular, and thyroid diseases or diabetes or cancer, did not take any supplements, vitamins or minerals in the past month or omega-3 supplements in the last three months, were included. Subjects who smoked or those who were on a speci c dietary pattern, were not included in the current study. Full explanations about the purposes and methods of the study were given to the participants and then their written consent was obtained.

Assessment of dietary intake
Dietary intake of participants was assessed using a 147-item, semi-quantitative Food Frequency Questionnaire (FFQ). The reliability and validity of this questionnaire has been con rmed previously (16).
Participants were asked to report the frequency of consumption of each food item per day, week, month, and year. These amounts were then converted to grams per day using household measurements. Finally, participants' energy, macro-and micro-nutrients intake were calculated using the Nutritionist 4 software, modi ed for Iranian foods.
Assessment of sleep quality Sleep quality was estimated with Epworth Sleepiness Scale (ESS), which has 8 questions (17). The subjects were asked to rate, on a 4-point scale (0-3), their probability of falling asleep while engaged in eight different activities. The ESS score was calculated by adding the scores for the eight questions. ESS scores were divided into two groups, zero to 10 as the desirable and 11-24 as the undesirable situation (18).

Assessment of other variables
Participants' physical activity was assessed using the International Physical Activity Questionnaire (IPAQ) (19), and converted to metabolic equivalents (MET) in terms of min/week (MET-min/week) based on references. The validity of Persian version of the questionnaire had been approved previously (20). The participants' weight was measured using Seca digital scale (755, Germany) to the nearest 100 gram with the subjects wearing light clothing without shoes. Participants' height was also measured by a stadiometer while the subject's shoulders were in normal position to the nearest 1 cm. BMI was calculated by dividing the weight (kg) to height squared (m 2 ). Eventually, participants' waist circumference was measured in the middle of last rib and iliac crest to the nearest 1 cm.

Statistical analysis
Factor analysis was used to identify major dietary patterns. First, due to the large number of food items, we assigned each food item into 22 prede ned food groups based on the similarity in their nutrient contents (Table 1). Principal component analysis was done with the factors rotated by orthogonal transformation. The natural interpretation of the factors in conjunction with eigenvalues (≥ 1.3) and Scree plot determined whether a factor should be retained. Identi ed dietary patterns were labelled on the basis of our interpretation of data and on prior literature. The factor score for each pattern was calculated by summing intakes of food groups weighted by their factor loadings, and each participant received a factor score for each identi ed pattern. Participants were then categorized based on tertiles of dietary pattern scores. To determine the association of dietary patterns with sleep quality, we used multivariate logistic regression. The rst tertile of dietary pattern scores was considered as the reference in each model. Regression analysis was controlled for participants' age, BMI, marital status, education, physical activity, and total energy intake in different models. All statistical analyses were done using the Statistical Package for Social Sciences software (SPSS Inc, version 16).     This study is the rst investigation about the "Relations of Dietary patterns and sleep disorders in shift workers" association of dietary patterns with risk of sleep disorders among shift workers in Iran. In a recent cross-sectional study among female university students, four major dietary patterns including mixed, high protein, western, and healthy were identi ed. Signi cant association was found only for the mixed dietary pattern in relation to the increased sleep quality (21). It seems that subjects on the mixed dietary pattern received su cient amounts of different micro-and macronutrients. Another study among schoolchildren found a signi cant inverse association between "dairy, bread, green leaves, maize/potatoes and sausages" dietary pattern and sleep duration (22). Low intake of fruit and vegetables and high levels of re ned carbohydrates were common among participants in that pattern.
Other studies also have shown that a higher diet quality is linked to a reduced risk of sleep disorders (23,24). Overall, it seems that consumption of high-quality diets, rich in fruits and vegetables, is linked to a reduced risk of SDs, while, adherence to a dietary patterns with high amounts of saturated fatty acids and animal food items and low amounts of fruit and vegetables is linked to an elevated risk of SDs (25)(26)(27).
The exact mechanisms through which dietary intakes is associated with sleep quality are unknown. Consumption of a high-carbohydrate and low-protein diet has been associated to the best quality of sleep. Tryptophan, a precursor of serotonin (28), enters the brain in a competitive manner with large-chain neutral amino acids (LCNAAs) (29). Dietary carbohydrates, unlike proteins, increase brain tryptophan concentrations in the circulation (30). Tryptophan enters the brain and up-regulates serotonin production (31), which then promotes sleeping (32). In contrast, diets high in solid oils and saturated fatty acids and low in bers may produce more nighttime arousals, and reduce overall sleep quality (21). Diets high in fats commonly have low amounts of carbohydrates. In addition, fat intake stimulates postprandial release of cholecystokinin, a satiety hormone released by the duodenum (33). Injection of cholecystokinin into rats has resulted in sleep promotion (34). Moreover, cholecystokinin was reported to be positively correlated with fatigue (35). Furthermore, meal time is another important factor that in uences on the association of dietary intakes with sleep quality (21). Nocturnal eating, in particular consumption of a high-fat diet, is associated to the reduced overall quality of sleep (36). Therefore, night shift workers are susceptible to sleep disorders caused by nocturnal eating which becomes more serious by consumption of a high-fat and low fruit and vegetables diet.
This is the rst study on the association of dietary patterns with sleep quality among shift workers. Adjustment for a list of probable confounding factors is a strength of the current study. However, limitations of the study also should be taken into account. This study had a cross-sectional design, therefore, it is impossible to confer the causality. In addition, a limited number of study participants had dietary data. A validated FFQ was used to assess dietary intakes of participants, however, misclassi cation of participants by using this assessment should not be neglected. Finally, sleep quality was measured by using a questionnaire in our study. Hence, self-reported information might be over-or under-estimated depending on the subject's memory.
In conclusion, current study showed a signi cant direct association between a dietary pattern high in animal proteins and solid oils and low in fruit and vegetables with risk of sleep disorders among shift workers in a cement factory. However, we failed to nd such an association for two other dietary patterns.
Further studies with large sample sizes are required to shed light in this issue.

Declarations
Ethics approval and consent to participate: This research has been approved by the Research Ethics Committee of Tehran University of Medical Sciences and Health Services, Tehran, Iran. Full explanations about the purposes and methods of the study were given to the participants and then their written consent was obtained.
Consent for publication: The authors would like to advise that all authors listed have contributed to the work and approved the content of the submitted manuscript. All authors have agreed to submit the manuscript to the "annals of general psychiatry".
Availability of data and material: All data generated or analysed during this study are included in this Competing interests: There is no con ict of interest in this study to declare.
Funding: This research has been supported by the Research Ethics Committee of Tehran University of Medical Sciences and Health Services, Tehran, Iran.Grant no. 30331-161-03-94.
Authors' contributions: Gity Sotoudeh conceived and developed the idea for the paper and revised the manuscript; Reihaneh Khorasaniha contributed to data collection and wrote numerous drafts; Alireza Milajerdi and Farahnaz Khajehnasiri contributed to data analysis and interpretation of the data.