Effects of Sleep Deprivation on Weight Loss and Gut Microbiota Diversity in Obese Patients on a Calorie Restrict Diet

Objective: This study aimed to investigate the effects of sleep deprivation (SD) on the weight loss and gut microbiota diversity in obese patients on a calorie restrict diet (CRD). Methods: Twenty obese patients who were divided into two groups: sleep deprivation group(SD group,n=10) and non-sleep deprivation group(NSD group ,n=10). All the patients received CRD for twelve weeks. Measurement of anthropometric parameters, biochemical examinations and gut microbiota detection were done at baseline and the end of week 12. MI Bands were used to monitor the sleep and exercise. Body shape parameters were measured by using the JAWON ioi353. The gut microbiota was examined by PCR in the v3-v5 region of 16S rDNA gene, and high-throughput sequencing was carried out on the Illumina Miseq platform. The operational taxonomic units (OTUs) was used for analysis. Results: 1. CRD improved the body weight (BW) , waist circumference(WC), blood pressure (BP),basal metabolic rate (BMR) ,body fat content(BFC),and Insulin resistance index (HOMA-IR) in all obese patients. 2. In the NSD group, the BW, BFC, VFA, BMR and total cholesterol (TC) reduced signicantly when compared with the NSD group after CRD intervention (P<0.05). 3. The Alpha diversity of gut microbiota remained unchanged after the intervention in two groups after CRD intervention. 4. There was a negative correlation between Mollicutes and BMR in the NSD group. Conclusion: The effects of CRD may be weaken by SD in weight loss and the metabolism of blood lipid. Mollicutes bacteria may be related to the weight loss after CRD intervention in obese patients.


Introduction
Obesity has been a common disease, threatening the physical and mental health in about 39% of people worldwide 1 . It is also a risk factor for many diseases such as cancers, cardiovascular diseases, endocrine system diseases and digestive system diseases. Safe and effective weight loss can reduce the risk of obesity-related diseases and even decrease all-cause mortality of obese patients 2 . Currently, many strategies have been developed to treat obesity such as dietary intervention, exercise, drugs and surgery, and dietary intervention is the rst step in the process of obesity management 3 . The calorie restrict diet (CRD) recommended by Chinese Nutrition Society is a nutritionally-balanced, safe and feasible strategy with favorable weight losing effect 4 . However, the weight losing effect varies among individuals 5 , but the speci c mechanism is still unknown.
Sleep deprivation (SD) refers to the forced reduction of sleep duration (generally <6 h within 24 h) due to various reasons 6 . Sleep deprivation can cause a series of physiological and psychological changes, leading to a variety of adverse consequences such as cognitive decline, metabolic disorders, and nally obesity 7 . In addition, SD may also affect the diversity of gut microbiota, which may compromise the weight loss [8][9] . Therefore, this study was undertaken to investigate the effects of SD on weight loss and gut microbiota diversity in obese patients on a CRD.

Subjects
Twenty subjects with simple obesity were recruited from Huai'an Second People's Hospital between September 2018 to December 2019.

Ethics statement
This study was approved by the Ethics Committee of Huai'an Hospital, Xuzhou Medical University and Huai'an Second People's Hospital (Huai'an city, Jiangsu Province, China).
All subjects signed the informed consent before study. This study was registered in the Clinical Trial Registration Center (No: ChiCTR1800015171).
Inclusion criteria were as follows: (1) Subjects were older than 18 years, but younger than 50 years; (2) the body mass index (BMI) was ≥ 28.0 kg/m 2 ; (3) All the subjects received the psychological assessments and CRD intervention.
Exclusion criteria were as follows: (1). Subjects were diagnosed with secondary obesity; (2) Subjects had gastrointestinal diseases or malabsorption; 3. Subjects were previously diagnosed with severe liver, kidney, heart or lung dysfunction,, hypothyroidism or malignant tumor; (4) Subjects were treated with hypoglycemic drugs, lipid-lowering drugs, weight loss drugs, glucocorticoids and other drugs affecting metabolism within the last 4 weeks; (5) Subjects received weight-loss surgery before study; (6). Subjects were previously diagnosed with mental disorders, and the inability to provide self-reported data and informed consent; (7). Subjects were pregnant or breast-feeding; (8) Subjects had history of alcohol abuse (amount of alcohol consumed per week was ≥140 g for men and ≥70 g for women); (9). Subjects were previously diagnosed physical dysfunction; (10) Subjects were participated in other weight loss programs or the body weight decreased by more than 5% in the past 6 months; (11) Subjects had poorly controlled severe hypertension (systolic blood pressure [SBP] ≥140mmHg, diastolic blood pressure [DBP] ≥90 mmHg).
Grouping: All the subjects were divided into two groups: SD group (n=10) and NSD group(n=10) . The daily sleep time was less than 6 h in the SD group, but longer than 7 h and shorter than 9 h in the NSD group 10 . After 12-week follow-up, 2 subjects one in SD group anther in NSD group. failed to complete the intervention and 2 one in SD group anther in NSD group. were treated with lipid-regulating and uricase-lowering drugs or experienced a tra c accident. Thus, there were excluded from this study. Finally, sixteen subjects (NSD group, n=8; SD group,n=8) completed this study. There were no signi cant differences in the gender, age, body weight (BW), body mass index (BMI) and biochemical parameters between two groups at baseline.

General information collection
After the questionnaire survey and psychological assessment, all subjects received CRD combined with appropriate exercise (at least 10,000 steps per day) for weight loss for 12 weeks.
In the CRD intervention, the daily energy intake was 20-25 kcal/kg per unit standard weight 11 . The standard weight (kg) = height (cm) -105. The proportion of energy at breakfast, lunch and dinner was 30%, 40% and 30%, respectively. The proportion of carbohydrate, protein and fat in the diet was 40%, 20% and 40% respectively. Meal replacement bars (Guangzhou NanDaFeiTe Nutrition Health Consulting Co. Ltd) were taken twice daily (before breakfast and before dinner). During the study, the MI Band was used to monitor the sleep and activities (the results of the MI Band met the requirements).
Body shape assessment, biochemical examination, liver color ultrasonography and gut microbiota detection were performed at baseline and 12 weeks after CRD intervention.

Questionnaire
The demographic characteristics, lifestyle information and medical history were collected by the trained investigators through a standard questionnaire at baseline. Current smokers were defined as subjects smoking at least one cigarette per day over the past months. Drinking was defined as drinking at least once weekly for nearly 1 month.
The mental health status was assessed with the Mental Healthy Assessment Scale (Symptom Checklist 90, SCL-90) 12 .

Data collection
Body shape parameters included basal metabolic rate (BMR), body fat content (BFC), body fat percentage (BFP), visceral fat area (VFA) and BW were measured with the Korean JAWON body composition analyzer ioi353.
The BMI was calculated as the weight (kg) divided by the height squared (m 2 ). Waist circumference (WC) was measured with a non-stretchable tape over the unclothed abdomen at the narrowest point between the lowest rib and the iliac crest. Two measures were obtained and the mean (centimeters) was calculated for analyses. The blood pressure (BP) was consecutively measured thrice (OMRON Model HEM-752 FUZZY, Omron Company, Dalian, Liaoning, China), and the mean was calculated for further use.
The estimated glomerular ltration rate (eGFR) was calculated based on the serum creatinine level by using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation 14 . Serum creatinine, 1 mg/dl = 88.41 μmol/l.

Hepatic ultrasonography
Hepatic ultrasonography was performed in all subjects by an experienced clinician using an Aloka Prosound α6 probe (Hitachi Aloka, Japan) with a 3-MHz probe in the same working conditions and operating standards. Before the ultrasonography, subjects experienced fasting for more than 8 h, and lied in a left-lateral position, and ultrasonography was done during breath holding (about 5s) at the end of expiration. The mild, moderate and severe fatty liver disease was determined according to the Doppler spectrum of hepatic vein pulse and liver hemodynamics 15 .

6 Detection of Gut microbiota
The feces (as big as the size of soybean) (about 2 g) were collected in the morning from subjects after fasting, stored separately in a sterile container at -80℃ at baseline and after 12-week CRD intervention. The complete genomic DNA was extracted and puri ed using the QIAamp DNA Stool Mini Kit (QIAGEN, Hilden, Germany) according to the manufacturer's instructions. Two-step PCR ampli cation was done to construct gene library. The v3-v5 region of bacterial 16S rDNA gene was ampli ed by PCR, high-throughput sequencing was carried out on the Illumina Miseq platform, and then the operational taxonomic unit (OTU) was analyzed and annotated with a bioinformatics software. After optimizing the gene sequence with Trimmomatic software, OTU clustering was carried out and compared with silva 123 database for species information annotation. Statistical analysis was conducted based on taxonomic information.

Statistical analysis
All statistical analyses were performed using SPSS version 22.0 (Statistical Product and Service Solutions Inc, Chicago, IL, USA). The quantitative data are expressed as means ± standard deviations; qualitative data are expressed as numbers or percentages. Paired t test was used to compare the data before and after intervention in the same group. Comparisons between groups were done using Student's t test for quantitative data and the Chi-square test for qualitative data. Stamp software was used to analyze the differences of Gut microbiota distribution. Kruskal-Wallis test was used to compare the differences in intestinal bacteria between groups, and Spearman correlation analysis was used to analyze the correlation between clinical indicators and host ora. SD 0 and NSD 0 represented the data at baseline in the SD group and NSD group, respectively. SD 12 and NSD 12 represented the data at the end of 12-week CRD intervention in the SD group and NSD group, respectively. A value of P<0.05 was considered signi cant statistically.

Characteristics of all subjects before and after 12-week CRD intervention
The WC, SBP, DBP, BFC, BW, BMR, HOMA-IR, TG and ALT,in all subjects signi cantly decreased after CRD intervention, as compared to those at baseline (Table 1). Table 1. Characteristics of all subjects before and after 12-week CRD intervention.

Variates
Before CRD intervention After CRD intervention P Characteristics of subjects in two groups before and after CRD intervention At baseline, there were no marked differences in the age, gender, BMI, WC, BW, SBP, DBP, BFC, BFP, BMR, VFA, HbA1c, HOMA-RI, HbA1c, TC, TG, ALT, AST, ALP, BUN and eGFR between SD group and NSD group (Table 2). After CRD intervention, the BW, WC, BMR, SBP and DBP decreased signi cantly in both groups. However, there were slight declines in the TG in the SD group and TC,HOMA-RI in the NSD group (Table 2). Notes: All variables are expressed as n (%) for qualitative data or mean ± standard deviation for quantitative data. Pa, the difference of the change of each parameter before and after intervention in the SD group. Pb, the difference of the change of each parameter before and after intervention in the NSD group. P0, the difference of baseline parameters between SD group  difference between baseline level and 12-week level.

Findings from Doppler ultrasonography
At baseline, fatty liver disease was noted in all the subjects of both SD group and NSD group. After 12-week intervention, fatty liver disease was not observed in 87.5% and 37.5% of subjects in the NSD group and SD group, respectively.
Gut microbiota distribution Sequencing depth of gut microbiota: the species accumulation curve of samples was analyzed by random sampling method to detect the sequencing depth. The results showed that the dilution curve tended to atten when there were 30 sequencing samples, indicating that the depth of this sequencing met the requirements, and the additional amount of sequencing data was not necessary for studying the discovery of OTU ( Figure 1).

Gut microbiota in two groups before and after CRD intervention
As shown in Fig.2A, the numbers of OTUs were similar between SD 0 and SD 12 , NSD 0 and NSD 12 , SD 0 and NSD 0 , and between SD 12 and NSD 12 (P SD0-SD12 =0.57, P NSD0-NSD12 =0.08, P SD0-NSD0 =0.29 and P SD12-NSD12 =0.43). However, as shown in Fig.2B, the abundance in the NSD group increased gradually after intervention, while that in the SD group decreased gradually, and the abundance in the SD group was lower than in the NSD group before and after CRD intervention,without signi cance.
Heat map analysis of correlation between gut microbiota abundance and clinical factors before and after intervention in two groups At the phylum level, heat maps were plotted based on the correlation coe cient between abundance of gut microbiota and clinical indicators. As shown in Figure 3, at baseline, in the NSD group, Selenomonadales was negatively related to AST, Bi dobacteriaceae and Cr. Coriobacteriaceae were negatively related to FBG. Desulfovibrionaceae was positively related to BFR. There was a negative correlation between Mollicutes and BMR.
However, there was no signi cant association in the SD group between abundance of gut microbiota and clinical factors.
At the end of 12 weeks after CRD intervention (Fig.4), in the NSD group, there was a signi cant positive correlation between Akkermansia and FBG. Bi dobacteriaceae was positively related to FBG, HbA1C and TG. Clostridiales was negatively related to BUN. There was a negative correlation between Mollicutes and BMR. However, in the SD group, Burkholderiales was positively related to FBG. Desulfovibrionaceae was positively related to AU. Actinobacteria and Actinomycetaceae were negatively related to VFA.
Mollicutes Alpha diversity in two groups before and after intervention.
A stable negative correlation between Mollicutes bacteria and BMR was noted in the NSD group before and after CRD intervention (Fig3-4). To better analyze the relations between Mollicutes and BMR, Mollicutes Alpha diversity was analyzed separately in the Figure 5. There were no signi cant difference in the relative abundance of Mollicutes before and after intervention in the SD group and NSD group (P SD0-SD12 =0.685, P NSD0-NSD12 =0.524) , However, the change of relative abundance of Mollicutes in the NSD group after intervention was signi cantly greater than in the SD group.

Discussion
It is estimated that more than 44% of the global population have overweight, and more than 300 million people are affected by obesity. In the past decade, the incidence of obesity has increased at an alarming rate 16 . Obesity has been a risk factor for many diseases, including type 2 diabetes mellitus (T2DM), nonalcoholic fatty liver disease and cardiovascular disease 17 . Therefore, obesity has been a global public health problem that should be managed actively. Our study showed, after 12 weeks of CRD and exercise intervention, some clinical parameters including WC, BW, SBP, DBP, BFC, BMR, HOMA-IR, TG and ALT signi cantly decreased. It suggests that CRD intervention can effectively reduce the BW and BFC, improve fat metabolism and relieve insulin resistance even in the presence of SD. Although this has been reported previously [18][19] , ndings in previous studies were from obese patients with T2DM, and there were no prospective studies. In our study, all subjects were diagnosed with simple obesity and CRD program was designed by a professional nutritionist and sports manager before study. Therefore, our results may be helpful to improve the management of obesity.
Studies have indicated that SD may promote the occurrence and development of obesity in children and adults [20][21] , however the relationship between SD and weight loss is poorly understood. Our study showed that CRD intervention had a good effect on the weight loss in all subjects and improved WC, BW, BMR,SBP and DBP . When the subjects were divided into SD and NSD group, results showed the BFC and TC in the NSD group signi cantly reduced after intervention, while both remained unchanged in the SD group. In addition, the changes of VFA, BW, BFC, BMR and TC in the NSD group were more obvious as compared to those in the SD group after CRD intervention. This further indicates that SD compromises the protective effect of CRD on the weight loss and lipid metabolism. Moreover, the prevalence of fatty liver disease in the NSD group and SD group reduced after weight loss following 12-week intervention, 87.5% of subjects in the NSD group had no fatty liver disease, which was signi cantly higher than in the SD group (37.5%). Therefore, SD may also compromise the improvement of fatty liver disease during the CRD intervention. However, the speci c mechanism is unclear. There are several possible explanations, (1) SD may lead to uncontrollable behaviors and poor diet compliance in the process of weight loss 20, 22 ; (2) sleep deprivation may decrease leptin, increase ghrelin as well as hunger and appetite, which all cause a decrease in the basal metabolic rate and an increase in the intake, ultimately leading to obesity 23 ; (3) SD may damage the compositions of gut microbiota 24 , thereby increasing intestinal permeability, compromising the metabolism of nutrients such as fat and sugar and nally resulting in low energy consumption and ultimately obesity 25 . Among these mechanisms, the most remarkable one is that SD damages the diversity of gut microbiota and thus compromises the protective effects of CRD.
To further investigate the potential mechanism, the gut microbiota was detected before and after CRD intervention in all subjects. 32 samples were examined before and after intervention, and the dilution curve results showed, if the number of samples was greater than 30, the measured data volume was reasonable. Thus, the number of samples in this study met the requirements. Studies have reported that diet can quickly and effectively change the compositions of gut microbiota 26-27 . However, no signi cant differences were found in the alpha diversity analysis of intergroup differences in gut microbiota ( Fig.2A). However, the abundance of the curve of gut microbiota at baseline and the end of 12 weeks after intervention changed to a certain extent. Results showed that, after 12-week intervention, the abundance of gut microbiota in the NSD group was the highest, and that in the SD group was almost no change. Another study also shows that SD had a certain effect on the abundance of gut microbiota 28 , which may affect the regulatory effect of diet on the gut microbiota. The reason why the gut microbiota abundance of the NSD group after CRD intervention was higher than that of the SD group after intervention might be ascribed to that SD may gradually compromise the rapid regulatory effect of diet on the gut microbiota.
In addition, the relationships between gut microbiota and clinical factors were evaluated in all the subjects. There were close correlations between species abundance of gut microbiota and clinical factors, and the diversity of gut microbiota in the SD group was inhibited at baseline and had no relationship with clinical factors. Most of these correlations were unstable and special, but a stable negative correlation between Mollicutes bacteria and BMR was noted in the NSD group before and after CRD intervention. Compared with the SD group, the relative abundance of Mollicutes increased signi cantly in the NSD group. However, there was no similar stable correlation between BMR and abundance of Mollicutes in the SD group.
Therefore, we speculate that SD disturbs the dynamic balance of gut microbiota,which leads to the stable correlation between BMR and abundance of Mollicutes to be broken. A study has shown that the recovery of low BMR is a sign of better weight loss 29 . Therefore, in obese patients without SD, increasing the abundance of Mollicutes may be bene cial to reduce BMR and improve the weight loss effect of CRD. However, the speci c mechanism should be elucidated in future studies.
Our study showed that CRD intervention reduced SBP and DBP in obese subjects with or without SD, but the change of DBP in the NSD group was more obvious than that in the SD group. A Study suggests that SD compromises the antihypertensive effect of CRD due to the SD induced increase of BP 30 . Interestingly, the change of SBP in the SD group was more obvious than in the NSD group. This might be ascribed to the higher SBP at baseline in the SD group and thus the BP was sensitive to dietary intervention, but this should be further con rmed in more studies.

Conclusions
In conclusion, our results indicate that CRD intervention may effectively reduce the BW, WC, BMR, SBP and DBP of obese subjects and also improve lipid metabolism in these subjects with normal sleep rhythms. SD may impair the weight loss effects of CRD intervention and inhibit the diversity of gut microbiota in simple obese subjects. Our study for the rst time investigates the relationships among SD, CRD intervention and gut microbiota. Our ndings suggest that Mollicutes bacteria may be an important factor affecting the weight loss effect of diet intervention in obese patients with SD. A further study on the Mollicutes may provide an important basis for the treatment of obesity by targeting gut microbiota in the future.
Unfortunately, there were limitations in our study,(1) this was a single center, prospective clinical study with small sample size.
(2) the CRD intervention was administered for only 12 weeks, which was relatively short. Therefore, more studies with large sample size are needed to con rm our ndings and investigate the effects of long term CRD intervention in obese subjects, which may provide a new strategy for the clinical management of obesity. Helsinki Declaration and its later amendments or comparable ethical standards.

Consent for publication
Consent to publish has been obtained from the participant to report individual patient data.

Availability of data and materials
The datasets during and/or analysed during the current study available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.  A. Gut microbiota abundance Alpha diversity between SD group and NSD group before and after intervention. B. Curve analysis of gut microbiota and species Rank-abundance.
Note: SD 0 : baseline in the SD group, SD 12 : the end of 12 weeks after CRD intervention in the SD group, NSD 0 : baseline in NSD group, NSD 12 : the end of 12 weeks after CRD intervention in the NSD group.