Effect of gut microbiota modulation on sleep: a systematic review and meta-analysis of clinical trials

1 ABSTRACT 1 Context: A bidirectional relationship between gut microbiota (GM) and circadian 2 rhythms has been proposed. Objective: The aim of this study was to analyse the efficacy 3 of probiotic or prebiotic intervention on sleep quality and quantity. Data sources, 4 selection and extraction: A systematic review and meta-analysis were conducted using 5 the databases PubMed-Medline, Embase, CINAHL and Web of Science. Only 6 randomised clinical trials written in English or Spanish were considered. The initial 7 search resulted in 219 articles. Following the removal of duplicates and consideration of 8 the selection criteria, 25 articles were selected for the systematic review and 18 articles 9 for the meta-analysis. Results: The effect of microbiota modulation showed no 10 significant improvement on sleep quality in our meta-analysis (p=0.31). In terms of sleep 11 duration, the meta-analysis found no improvement due to GM modulation (p=0.43). 12 Conclusions: Our meta-analysis results indicate that there is still insufficient evidence to 13 support the relationship between GM modulation and improved sleep quality. While 14 several studies assume that including probiotics in the diet will undoubtedly improve 15 sleep quality, more research is needed to fully understand this phenomenom. Systematic 16 Review Registration: PROSPERO registration no. CRD42021245118. 17 18 2 INTRODUCTION 19 Gut microbiota (GM) refers to the set of more than 1000 species of living microorganisms 20 such as viruses, bacteria, archaea, fungi and protists that colonise the gastrointestinal 21 tract.1 The GM constitutes a host-specific ecosystem that is genetically inherited and 22 matures during childhood, affecting individuals’ physiological homeostasis.2 The 23 functions of GM include controlling the proliferation of pathogenic bacteria, regulating 24 nutrient absorption and metabolism, modulating the immune system and participating in 25 the synthesis of metabolites such as short-chain fatty acids (SCFA) and vitamins that 26 provide benefits to the host.3–6 27 GM composition depends on several factors such as diet, ingestion of certain drugs, age 28 and genetics, but in general the dominant microbial communities are the phyla 29 Firmicutes, Bacteroidetes, Actinobacteria and Proteobacteria.7 Certain factors can affect 30 the diversity of GM leading to a process known as intestinal dysbiosis.8,9 It has been 31 proposed that the phenomenon of dysbiosis could be related to the development of 32 disorders, such as obesity and autism, through what is known as the gut-brain axis 33 (GBA).10–12 The gut has an intrinsic innervation, the enteric nervous system (ENS), 34 composed of sensory neurons that detect physicochemical changes, motor neurons and 35 interneurons located in enteric ganglia forming a complex network of more than 500 36 million neurons.13 In succession, the nervous cells of the ENS can communicate with 37 brain neurons and GM via GBA. This communication system is bidirectional and is 38 carried out in three main pathways: the vagus nerve, the systemic circulation and the 39 immune system involving neurotransmitters and neuromodulators.13 40 3 Due to the bidirectional relationship between the central nervous system (CNS) and GM, 41 several studies have associated sleep disturbances with changes in GM composition, as 42 they observed that patients with insomnia had lower microbial richness and diversity 43 caused by a depletion of anaerobic bacteria and SCFA-producing bacteria related to 44 control individuals.14,15 Lachnospira and Bacteroides, in particular, are common bacteria 45 in patients with acute insomnia, whereas Faecalibacterium and Blautia are frequent in 46 patients with chronic insomnia.14 47 Regarding sleep, it is worth noting that it plays a key role in development, growth and 48 immune function. Insufficient sleep has been associated with an increased risk of 49 mortality and the development of cardiovascular and coronary heart disease, hypertension 50 and diabetes.16 Additionally, it has been proposed that sleep quality disturbances may be 51 driven by hypothalamic-pituitary-adrenal (HPA) axis overactivity and cortisol 52 hypersecretion.17 The beneficial effects of GM on the CNS via the GBA have led some 53 researchers to suggest that GM interacts with the circadian system, as the gut influences 54 energy status by controlling physiological functions such as digestion and nutrient 55 absorption, functions that are also regulated by the genetic clock. 56 The use of probiotics and prebiotics to manage intestinal dysbiosis have been tested in 57 humans. For example, the administration of probiotics containing live microorganisms 58 showed beneficial effects on pathologies such as diarrhoea, colitis, and inflammatory 59 bowel disease.18,19 Prebiotics promote the selective growth of bacteria by preventing the 60 growth of pathogens in the colonic flora. The food components that seem to exert the best 61 prebiotic effect are fructooligosaccharides (FOS), inulin and lactulose.20,21 In recent 62 years, the administration of prebiotics and probiotics has been investigated with the aim 63 of treating insomnia in both experimental animals and humans, although contradictory 64 4 data have been found.22 For example, the administration of Lactobacillus brevis DL1-11, 65 L. brevis SBC8803 and L. fermentum PS150TM was able to increase sleep duration and 66 decrease sleep latency in mice.23–25 Similarly, treatments with Lactobacillus gasseri 67 CP2305 and Lactobacillus casei strain Shirota (LcS) improved sleep quality in healthy 68 adults affected by stress.26,27 However, another study found no changes in sleep quality 69 after administration of Lactobacillus acidophilus Rosell-52 and Bifidobacterium 70 longum Rosell-175 in stress-affected volunteers.28 Regarding prebiotics such as inulin or 71 galacto-oligosaccharides (GOS), it seems that they have no effect on sleep disturbances 72 either in adults or in children.29,30 73 The main hypothesis proposed in relation to the sedative effect of some probiotic strains 74 is associated to the immune system modulation. The production of pro-inflammatory 75 cytokines such as interleukin 1-beta (IL-1β) and tumour necrosis factor alpha (TNF-α) 76 has been found to be increased in rats exposed to chronic sleep restriction, indicating the 77 presence of inflammatory processes in sleep-restricted experimental animals.31 In this 78 regard, there is evidence of anti-inflammatory effects of probiotic strains such as 79 Lactobacillus plantarum, which hinders the expression of IL-1β and TNF-α while 80 increasing the expression of IL-10, an anti-inflammatory cytokine, in rats treated with 81 2,4,6-trinitrobenzenesulfonic acid (TNBS)-induced colitis.32 82 To date, evidence for the effect of probiotics on sleep quality is very limited,33,34 but no 83 evidence has been found for other modulators of GM, such as prebiotics or FMT. Thus, 84 the overall aim of this review was to analyse how GM intervention by probiotics or 85 prebiotics affects sleep quality and quantity, with other factors such as sleep duration as 86 secondary objectives. 87 5 METHODS 88 A systematic review was conducted with studies until September 2022 following the 89 Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) 90 guidelines35 (Table S1) and the Cochrane systematic review manual.36 The protocol for 91 this study was registered in PROSPERO (CRD42021245118). 92 Databases 93 The research question used in the search for information within this review followed the 94 Patient-Intervention-Outcomes (PIO) structure (Table 1): “How does GM modulation 95 influence sleep quality?”, as a consequence of the function of the intestinal microbiota 96 like an anti-inflammatory and neuroprotective agent of the central nervous system that 97 interacts with the circadian system. 98 [INSERT TABLE 1 AROUND HERE] 99 Four databases were used to collect data (PubMed-Medline, CINAHL, Embase and Web 100 of Science), and search strategies were developed by combining structured language 101 (Medical Subject Heading) and natural language. The strategies adapted to each database 102 are shown in Table S2. 103 Study eligibility criteria 104 The inclusion criteria used for this review were as follows: randomised clinical trials 105 (RCTs) performed in humans, published in Spanish or English, and relevant to the aim of 106 the study, including a measure of sleep quality and duration. Reviews, conference 107 abstracts, qualitative studies, or descriptive or observational studies without a control 108 6 group were excluded. Similarly, studies conducted on the paediatric population were 109 discarded. No studies were excluded due to methodological flaws or age restrictions. 110 Selection of studies and methodological quality 111 Two authors (EGH and CRG) carried out the article selection process independently; in 112 case of discrepancy, a third researcher (PR) was consulted. Likewise, the methodological 113 quality was evaluated using the Jadad scale for randomised clinical trials. This scale is 114 composed of 5 items that serve to evaluate the quality of RCTs in a simple manner.37 A 115 score of more than 3 points is associated with a high-quality study, while a score of less 116 than 3 points is associated with a low-quality study; these scores are based on the quality 117 of double blinding, randomization and follow-up.38 It should be noted that this scale 118 covers some of the elements related to biases, thus acquiring reliability and external 119 validity.39 120 Data extraction 121 Data were extracted independently by two investigators (EGH and CRG) from each of 122 the studies and tabulated as follows: (1) reference; (2) number of participants; (3) 123 intervention used; (4) duration of intervention; (5) variables analyzed; and (6) main 124 outcomes. 125 The WebPlotDigitizer program was used to obtain the data from those articles that 126 provided only the relevant results in graphs. This software has already been used by other 127 authors in meta-analyses, demonstrating its effectiveness in extracting data from 128 graphs.40,41 129 Statistical analysis 130 7 The meta-analysis of the included trials was performed using Review Manager (RevMan) 131 software version 5.4. 132 Effect size 133 The effect size was measured by the mean difference and standard deviation or standard 134 error between the control and experimental groups. 135 For the main variable, sleep quality, the following questionnaires were used: Pittsburgh 136 Sleep Quality Questionnaire (PSQI), Numeric Rating Scale (NRS) and Athens Insomnia 137 Scale (AIS). 138 The studies included in the meta-analysis had to provide data on means, standard 139 deviation or standard error; or could be calculated from other data provided. Otherwise, 140 the corresponding author was contacted and, if no response was received, they were 141 discarded from the meta-analysis. 142 Heterogeneity and publication bias 143 Depending on the degree of heterogeneity (I2) obtained, a random-effects model (I2≥50%) 144 or a fixed-effects model (I2 ≤50%) was used to perform the meta-analysis. A forest plot 145 represented the analysis. In addition, the p-value was considered significant when it was 146 <0.05. 147 A sensitivity analysis was also used to observe whether changes in the data or in the 148 objective of the analysis affected the results and conclusions of this review. If after 149 sensitivity analysis, the conclusions did not change, they were considered robust. For this 150 8 purpose, each of the studies was excluded independently to identify which of them had a 151 significant influence on the findings. 152 RESULTS 153 After searching the different databases, 219 articles were initially retrieved. One hunded 154 fifty-one articles were screened by reading their title and abstract after removing 155 duplicates (68 documents). The number of publications considered in this systematic 156 review was reduced to 25 after full-text analysis, with 18 trials included in the meta-157 analysis. The entire process is depicted in Figure 1. 158 [INSERT FIGURE 1 AROUND HERE] 159 Included studies 160 The study population included a higher proportion of healthy people and students (60%). 161 Overall, the studies incorporated a total of 2192 participants, with a mean of 91.33 162 individuals (max= 581, min= 10) and a mean intervention duration of 7.12 weeks (max= 163 24, min=1). 164 Most studies used probiotics as modulators of GM. The most-used probiotic species were 165 Lactobacillus and Bifidobacterium, though Saccharomyces cerevisiae, Streptococcus and 166 Enterococcus faecalis were also used, as were three types of prebiotics (Inulin, GOS and 167 FOS). Different strains of Lactobacillus (acidophilus, bulgaris, rhamnosus, casei, 168 plantarum, fermentum, delbrueckii ssp. Bulgaricus, reuteri, gasseri, casei Shirota, 169 helveticus) and of Bifidobacterium (animalis ssp. lactis, longum, adolescentis, breve, 170 bifidum, longum ssp. infantis) were used in the assays. Only one study used prebiotics, 171 primarily inulin and fructo-oligosaccharides, as GM modifiers. Furthermore, studies that 172 9 used synbiotics in the intervention group were included. Lactobacillus and 173 Bifidobacterium strains were used as probiotics, and inulin and FOS were used as 174 prebiotics. 175 Sleep quality was analysed in 88.46% of the selected studies, while sleep duration was 176 analysed in 30.77%. The results extracted from the RCTs, as well as their characteristics, 177 are shown in Table 2. 178 [INSERT TABLE 2 AROUND HERE] 179 Methodological quality: assessment of bias 180 Figure 2 shows the overall score for each item, with the lowest risk of bias in studies that 181 included randomisation and a description of participant blinding. However, the trials did 182 not usually state how this randomisation and blinding were performed, nor whether the 183 fate of the participants was known. On the other hand, Figure 3 shows the evaluation of 184 the methodological quality and risk of bias using the Jadad scale of the studies included 185 in the systematic review. With the exception of the studies by Lee et al. (2021), Quero et 186 al. (2021), Rode et al. (2022), Shafie et al. (2022), Valle et al. (2021), Wälivaara et al. 187 (2019) and Wang et al. (2021), whose methodological quality is quite high, this figure 188 shows that some studies have items with some deficiencies.26–29,42–62,52,59,60 189 [INSERT FIGURE 2 AND 3 AROUND HERE] 190 Sleep quality and duration 191 Treating healthy students, we found several articles.26,27,43,45,49,51,56,63 All of them used 192 probiotics as the intervention in their experimental groups, mostly using Lactobacillus as 193 16 Interestingly, recent findings suggest that immune system cells can influence GBA since 327 an inflammatory environment in the intestinal mucosa can increase permeability and thus 328 impact brain activity78, as some cytokines have been shown to cross the blood brain 329 barrier (BBB) and promote neuroinflammation.79 In terms of immune system status, 330 Wang et al. (2021) concluded that a combination of probiotic strains including 331 Bifidobacterium longum, Lactobacillus acidophilus, and Enterococcus faecalis were able 332 to reduce plasma IL-6 levels in older adults on the 5th-7th day post-surgery 44. This effect 333 on IL-6 was also shown in immature enterocytes utilising probiotic conditioned media 334 (PCM) from Bifidobacterium longum subsp infantis and Lactobacillus acidophilus, but 335 not isolated organisms, which attenuated and decreased IL-6 induced by IL-1.80,81 336 Conversely, IL-1Ra, IL-10, IL-1b, IL-6, IL-8, and TNF-α concentrations did not alter 337 after administration of Lactobacillus acidophilus (DDS-1), Bifidobacterium animalis spp. 338 lactis (UABla-12) and L. rhamnosus (JB-1) to the intervention group.50,60 In this context, 339 L. rhamnosus GG (LGG) has not been demonstrated to change IL-10 levels, but has been 340 reported to increase IL-10R2 receptor subunit expression and decrease TNFexpression 341 in the mouse colon.82 Another study found that LGG-conditioned medium reduced TNF342 production in LPS-activated macrophages.83 343 Cortisol is well known to influence the sleep quality and duration, as high levels increase 344 waking time and lack of sleep increases cortisol levels.84 The effect of probiotic 345 supplementation on salivary and plasma cortisol levels has yielded contradictory results. 346 In only three of the trials included in this study, blood and salivary cortisol levels were 347 altered after administration of Bifidobacterium, Lactobacillus, and Enterococcus 348 strains.44,49,60 In this manner, probiotic supplementation has been documented in certain 349 trials to significantly reduce plasma cortisol levels in stressed-out students and adults.85,86 350 17 In the other trials evaluated, probiotics failed to lower cortisol levels.27,43,50,63 This is 351 consistent with a current clinical trial in which Bifidobacterium breve CCFM1025 was 352 administered to patients with major depression, yet failed to lower serum cortisol levels.87 353 Numerous studies on microbiota-derived metabolites have suggested that low SCFA 354 concentrations may contribute to the development of neuropsychiatric diseases, obesity, 355 and autism.88–90 Additionally, greater faecal propionic acid levels in infants are associated 356 with longer periods of uninterrupted sleep.91 These findings show that consuming 357 probiotic strains derived from SCFA-producing bacteria could be an attractive strategy. 358 Only Nishida et al. (2019) studied the impacts of probiotic consumption on SCFAs, 359 concluding that Lactobacillus gasseri probiotic supplementation increased n-valeric 360 levels in healthy volunteers but did not affect other SCFAs such as propionic and acetic 361 acids.27 Similar, research using Bifidobacterium animalis subsp. lactis BB-12 (BB-12) 362 found that it did not significantly enhance faecal SCFA levels in healthy adults.92,93 363 Limitations 364 The number of studies included in the meta-analysis is limited owing to lack of data sets 365 required to process and complete the analysis. While we attempted to contact all 366 corresponding authors of studies that did not contain this information, we did not always 367 receive a response. As a result, four studies must be excluded from the meta-analysis and 368 the systematic review.42,46,49,51 Another limitation to consider when interpreting the 369 results of this meta-analysis is the large variability in intervention length, which ranges 370 from 1 week in the study by Wang et al. (2021) to 24 weeks in the study by Nishida et al. 371 (2019), therefore this heterogeneity may have an impact on the analysis and the 372 magnitude of outcomes between interventions may not be accurately examined. 373 18 Regarding the administration of prebiotics, only one RCT was fitted in the inclusion 374 criteria, therefore we cannot draw a precise conclusion about the influence of prebiotics 375 on sleep. Another aspect to consider when interpreting and analysing the results is that a 376 significant proportion of the included studies used self-report scales to assess sleep 377 quality. In addition to these validated scales95-97, it would have been interesting to include 378 objective measures of sleep quality, with the PSQI being the most commonly used in the 379 studies selected. Finally, since it was outside the scope of this manuscript, the response 380 of the gut microbiome to probiotic or prebiotic intervention was not assessed in this 381 systematic review and meta-analysis, and thus the effect of microbiome modulation from 382 the interventions could not be determined. Rather than concluding the issue, our findings 383 pave way for future research to determine which interventions are associated with a 384 measurable gut microbiome response and whether this correlates with changes in sleep 385 quality. Likewise, future clinical trials should consider detailing the blinding methods and 386 participant dropout rate in order to improve rigour and reproducibility, as these items rank 387 lowest in our quality assessment of included studies. 388 CONCLUSIONS 389 Although several studies agree that modulation of the microbiota affects sleep quality, 390 the results of our meta-analysis do not show that the implementation of probiotics in the 391 diet leads to a significant improvement in sleep quality. The vast majority of the bacteria 392 used in the studies included in this study were from the Lactobacillus and Bifidobacterium 393 families, while Enterococcus faecalis, Streptococcus thermophilus, and Saccharomyces 394 cerevisiae were also used. 395 19 Few studies provided sufficient data to be included in our meta-analysis in terms of sleep 396 duration, which did not include statistically significant data. Because of this, it would be 397 interesting to expand research in this area to have access to a larger amount of data and 398 therefore be able to obtain a more exhaustive and precise conclusion. Only one clinical 399 trial with prebiotics was included in this study, which used Inulin and FOS in its 400 experimental group. Due to limited availability of studies obtained during our search, we 401 are unable to draw any conclusions on the effects of including prebiotics in the diet; 402 hence, it would also be worth conducting additional research in this area. 403 ACKNOWLEDGMENTS 404 The CTS-451 Research Group and the CEINSA Research Centre have supported this 405 research. 406 Funding. This research received no specific grant from any funding agency in the public, 407 commercial, or not-for-profit sectors. 408 Declarations of interest. The authors have no relevant interests to declare. 409 SUPPORTING INFORMATION 410 Table S1: PRISMA checklist. 411 Table S2: Search strategies in each database used. 412 REFERENCES 413 1. Matijašić M, Meštrović T, Paljetak HČ, Perić M, Barešić A, Verbanac D. Gut 414 microbiota beyond bacteria—mycobiome, virome, archaeome, and eukaryotic 415 20 parasites in IBD. Int J Mol Sci. 2020;21(8). doi:10.3390/ijms21082668 416 2. Derrien M, Alvarez AS, de Vos WM. The gut microbiota in the first decade of life. 417 Trends Microbiol. 2019;27(12):997-1010. doi:10.1016/j.tim.2019.08.001 418 3. Cooke G, Behan J, Costello M. Newly identified vitamin K-producing bacteria 419 isolated from the neonatal faecal flora. Microb Ecol Health Dis. 2009;18(3-4):133-420 138. doi:10.1080/08910600601048894 421 4. Tanner SA, Chassard C, Rigozzi E, Lacroix C, Stevens MJA. Bifidobacterium 422 thermophilum RBL67 impacts on growth and virulence gene expression of 423 Salmonella enterica subsp. enterica serovar Typhimurium. BMC Microbiol. 424 2016;16(1). doi:10.1186/S12866-016-0659-x 425 5. Pryde SE, Duncan SH, Hold GL, Stewart CS, Flint HJ. The microbiology of 426 butyrate formation in the human colon. FEMS Microbiol Lett. 2002;217(2):133-427 139. doi:10.1111/j.1574-6968.2002.tb11467.x 428 6. Ménard S, Laharie D, Asensio C, et al. Bifidobacterium breve and Streptococcus 429 thermophilus secretion products enhance T helper 1 immune response and 430 intestinal barrier in mice. Exp Biol Med. 2005;230(10):749-756. 431 doi:10.1177/153537020523001008 432 7. Thursby E, Juge N. Introduction to the human gut microbiota. Biochem J. 433 2017;474(11):1823-1836. doi:10.1042/bcj20160510 434 8. Jeffery I., O’Toole P., Öhman L, et al. An irritable bowel syndrome subtype 435 defined by species-specific alterations in faecal microbiota. Gut. 2012;61(7):997-436 21 1006. doi:10.1136/gutjnl-2011-301501 437 9. Isaac S, Scher J, Djukovic A, et al. Shortand long-term effects of oral vancomycin 438 on the human intestinal microbiota. J Antimicrob Chemother. 2017;72(1):128-136. 439 doi:10.1093/jac/dkw383 440 10. Bermudez-Martin P, Becker JAJ, Caramello N, et al. The microbial metabolite p-441 Cresol induces autistic-like behaviors in mice by remodeling the gut microbiota. 442 Microbiome 2021 91. 2021;9(1):1-23. doi:10.1186/s40168-021-01103-z 443 11. Roman P, Rueda-Ruzafa L, Cardona D, Cortes-Rodríguez A. Gut–brain axis in the 444 executive function of austism spectrum disorder. Behav Pharmacol. 445 2018;29(7):654-663. doi:10.1097/fbp.0000000000000428 446 12. Tomé-Castro XM, Rodriguez-Arrastia M, Cardona D, Rueda-Ruzafa L, Molina-447 Torres G, Roman P. Probiotics as a therapeutic strategy in obesity and overweight: 448 a systematic review. Benef Microbescrobes. 2021;12(1):5-15. 449 doi:10.3920/bm2020.0111 450 13. Rao M, Gershon MD. The bowel and beyond: the enteric nervous system in 451 neurological disorders. Nat Rev Gastroenterol Hepatol. 2016;13(9):517. 452 doi:10.1038/nrgastro.2016.107 453 14. Li Y, Zhang B, Zhou Y, et al. Gut microbiota changes and their relationship with 454 inflammation in patients with acute and chronic insomnia. Nat Sci Sleep. 455 2020;12:895-905. doi:10.2147/nss.S271927 456 15. Matenchuk B, Mandhane P, Kozyrskyj A. Sleep, circadian rhythm, and gut 457 22 microbiota. Sleep Med Rev. 2020;53. doi:10.1016/j.smrv.2020.101340 458 16. Medic G, Wille M, Hemels ME. Shortand long-term health consequences of sleep 459 disruption. Nat Sci Sleep. 2017;9:151. doi:10.2147/nss.s134864 460 17. Roth T. Insomnia: Definition, prevalence, etiology, and consequences. J Clin Sleep 461 Med. 2007;3(5 Suppl):S7-S10. doi:10.5664/jcsm.26929 462 18. Wildt S, Munck L, Vinter-Jensen L, et al. Probiotic treatment of collagenous 463 colitis: a randomized, double-blind, placebo-controlled trial with Lactobacillus 464 acidophilus and Bifidobacterium animalis subsp. Lactis. Inflamm Bowel Dis. 465 2006;12(5):395-401. doi:10.1097/01.mib.0000218763.99334.49 466 19. Li B, Liang L, Deng H, Guo J, Shu H, Zhang L. Efficacy and safety of probiotics 467 in irritable bowel syndrome: a systematic review and meta-analysis. Front 468 Pharmacol. 2020;11:332. doi:10.3389/fphar.2020.00332 469 20. Micka A, Siepelmeyer A, Holz A, Theis S, Schön C. Effect of consumption of 470 chicory inulin on bowel function in healthy subjects with constipation: a 471 randomized, double-blind, placebo-controlled trial. Int J Food Sci Nutr. 472 2017;68(1):82-89. doi:10.1080/09637486.2016.1212819 473 21. Goto H, Takemura N, Ogasawara T, et al. Effects of fructo-oligosaccharide on 474 DSS-induced colitis differ in mice fed nonpurified and purified diets. J Nutr. 475 2010;140(12):2121-2127. doi:10.3945/jn.110.125948 476 22. Sen P, Molinero-Perez A, O’Riordan KJ, McCafferty CP, O’Halloran KD, Cryan 477 JF. Microbiota and sleep: awakening the gut feeling. Trends Mol Med. 478 23 2021;27(10):935-945. doi:10.1016/j.molmed.2021.07.004 479 23. Yu L, Han X, Cen S, et al. Beneficial effect of GABA-rich fermented milk on 480 insomnia involving regulation of gut microbiota. Microbiol Res. 2020;233:126409. 481 doi:10.1016/j.micres.2020.126409 482 24. Higo-Yamamoto S, Yamamoto S, Miyazaki K, et al. Dietary heat-killed 483 Lactobacillus brevis SBC8803 attenuates chronic sleep disorders induced by 484 psychophysiological stress in mice. J Nutr Sci Vitaminol (Tokyo). 2019;65(2):164-485 170. doi:10.1016/0197-2456(95)00134-4 486 25. Lin A, Shih CT, Huang CL, Wu CC, Lin CT, Tsai YC. Hypnotic effects of 487 Lactobacillus fermentum PS150TM on pentobarbital-induced sleep in mice. 488 Nutrients. 2019;11(10):2409. doi:10.3390/nu11102409 489 26. Takada M, Nishida K, Gondo Y, et al. Beneficial effects of Lactobacillus casei 490 strain Shirota on academic stress-induced sleep disturbance in healthy adults: a 491 double-blind, randomised, placebo-controlled trial. Benef Microbes. 492 2017;8(2):153-162. doi:10.3920/bm2016.0150 493 27. Nishida K, Sawada D, Kuwano Y, Tanaka H, Rokutan K. Health benefits of 494 Lactobacillus gasseri CP2305 tablets in young adults exposed to chronic stress: a 495 randomized, double-blind, placebo-controlled study. Nutrients. 2019;11(8):1859. 496 doi:10.3390/nu11081859 497 28. Diop L, Guillou S, Durand H. Probiotic food supplement reduces stress-induced 498 gastrointestinal symptoms in volunteers: a double-blind, placebo-controlled, 499 randomized trial. Nutr Res. 2008;28(1):1-5. doi:10.1016/j.nutres.2007.10.001 500 24 29. Buigues C, Fernández-Garrido J, Pruimboom L, et al. Effect of a prebiotic 501 formulation on frailty syndrome: a randomized, double-blind clinical trial. Int J 502 Mol Sci. 2016;17(6):932. doi:10.3390/ijms17060932 503 30. Capitão LP, Baião R, Baek HK, et al. Prebiotic supplementation does not affect 504 reading and cognitive performance in children: a randomised placebo-controlled 505 study. J Psychopharmacol. 2020;34(1):148-152. doi:10.1177/0269881119862534 506 31. Zielinski MR, Kim Y, Karpova SA, McCarley RW, Strecker RE, Gerashchenko 507 D. Chronic sleep restriction elevates brain interleukin-1 beta and tumor necrosis 508 factor-alpha and attenuates brain-derived neurotrophic factor expression. Neurosci 509 Lett. 2014;580:27-31. doi:10.1016/j.neulet.2014.07.043 510 32. Satish Kumar CSV, Kondal Reddy K, Reddy AG, et al. Protective effect of 511 Lactobacillus plantarum 21, a probiotic on trinitrobenzenesulfonic acid-induced 512 ulcerative colitis in rats. Int Immunopharmacol. 2015;25(2):504-510. 513 doi:10.1016/j.intimp.2015.02.026 514 33. Irwin C, McCartney D, Desbrow B, Khalesi S. Effects of probiotics and 515 paraprobiotics on subjective and objective sleep metrics: a systematic review and 516 meta-analysis. Eur J Clin Nutr 2020 7411. 2020;74(11):1536-1549. 517 doi:10.1038/s41430-020-0656-x 518 34. Pacheco AP, Cedernaes J, Benedict C. Gut microbiome as a therapeutic target in 519 the treatment of sleep disorders: where we are. Sleep Med Rev. 2021;60:101547. 520 doi:10.1016/j.smrv.2021.101547 521 35. Page MJ, McKenzie JE, Bossuyt PM, et al. The PRISMA 2020 statement: an 522 25 updated guideline for reporting systematic reviews. BMJ. 2021;372. 523 doi:10.1136/bmj.n71 524 36. Higgins JPT, Altman DG, Gøtzsche PC, et al. The Cochrane Collaboration’s tool 525 for assessing risk of bias in randomised trials. BMJ. 2011;343(7829). 526 doi:10.1136/bmj.d5928 527 37. Clark HD, Wells GA, Huët C, et al. Assessing the quality of randomized trials: 528 reliability of the Jadad scale. Control Clin Trials. 1999;20(5):448-452. 529 doi:10.1016/s0197-2456(99)00026-4 530 38. Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of 531 randomized clinical trials: is blinding necessary? Control Clin Trials. 532 1996;17(1):1-12. doi:10.1016/0197-2456(95)00134-4 533 39. Halpern SH, Douglas MJ. Appendix: Jadad scale for reporting randomized 534 controlled trials. Evidence-based Obstet Anesth Oxford, UK Blackwell Publ Ltd. 535 Published online 2005:237-238. doi:10.1002/9780470988343.app1 536 40. Guyot P, Ades A, Ouwens M, Welton N. Enhanced secondary analysis of survival 537 data: reconstructing the data from published Kaplan-Meier survival curves. BMC 538 Med Res Methodol. 2012;12(1):1-13. doi:10.1186/1471-2288-12-9 539 41. Tsafnat G, Glasziou P, Choong M, Dunn A, Galgani F, Coiera E. Systematic 540 review automation technologies. Syst Rev. 2014;3(1). doi:10.1186/2046-4053-3-541 74 542 42. Harnett JE, Pyne DB, McKune AJ, Penm J, Pumpa KL. Probiotic supplementation 543 32 in increased intestinal permeability induced by central nervous system injury. CNS 675 Neurosci Ther. 2020;26(8):783-790. doi:10.1111/cns.13401 676 79. Banks WA, Farr SA, La Scola ME, Morley JE. Intravenous human interleukin-1α 677 impairs memory processing in mice: Dependence on blood-brain barrier transport 678 into posterior division of the septum. J Pharmacol Exp Ther. 2001;299(2):536-679 541. 680 80. Meng D, Zhu W, Ganguli K, Shi HN, Walker WA. Anti-inflammatory effects of 681 Bifidobacterium longum subsp infantis secretions on fetal human enterocytes are 682 mediated by TLR-4 receptors. Am J Physiol - Gastrointest Liver Physiol. 683 2016;311(4):G744. doi:10.1152/ajpgi.00090.2016 684 81. Guo S, Guo Y, Ergun A, Lu L, Walker WA, Ganguli K. Secreted metabolites of 685 Bifidobacterium infantis and Lactobacillus acidophilus protect immature human 686 enterocytes from IL-1β-induced inflammation: A transcription profiling analysis. 687 PLoS One. 2015;10(4):e0124549. doi:10.1371/journal.pone.0124549 688 82. Mirpuri J, Sotnikov I, Myers L, et al. Lactobacillus rhamnosus (LGG) regulates 689 IL-10 signaling in the developing murine colon through upregulation of the IL-690 10R2 receptor subunit. PLoS One. 2012;7(12). doi:10.1371/journal.pone.0051955 691 83. Peña JA, Versalovic J. Lactobacillus rhamnosus GG decreases TNF-α production 692 in lipopolysaccharide-activated murine macrophages by a contact-independent 693 mechanism. Cell Microbiol. 2003;5(4):277-285. doi:10.1046/j.1462-694 5822.2003.t01-1-00275.x 695 84. Hirotsu C, Tufik S, Andersen ML. Interactions between sleep, stress, and 696 33 metabolism: From physiological to pathological conditions. Sleep Sci. 697 2015;8(3):143-152. doi:10.1016/j.slsci.2015.09.002 698 85. Venkataraman R, Madempudi RS, Neelamraju J, et al. Effect of multi-strain 699 probiotic formulation on students facing examination stress: A double-blind, 700 placebo-controlled study. Probiotics Antimicrob Proteins. 2021;13(1):12-18. 701 doi:10.1007/s12602-020-09681-4 702 86. Chong HX, Yusoff NAA, Hor YY, et al. Lactobacillus plantarum DR7 alleviates 703 stress and anxiety in adults: a randomised, double-blind, placebo-controlled study. 704 Benef Microbes. 2019;10(4):355-373. doi:10.3920/bm2018.0135 705 87. Tian P, Chen Y, Zhu H, et al. Bifidobacterium breve CCFM1025 attenuates major 706 depression disorder via regulating gut microbiome and tryptophan metabolism: A 707 randomized clinical trial. Brain Behav Immun. 2022;100:233-241. 708 doi:10.1016/j.bbi.2021.11.023 709 88. MacFabe, Cain, RodriguezCapote K, et al. Neurobiological effects of 710 intraventricular propionic acid in rats: Possible role of short chain fatty acids on 711 the pathogenesis and characteristics of autism spectrum disorders. Behav Brain 712 Res. 2007;176(1):149-169. doi:10.1016/j.bbr.2006.07.025 713 89. Zhang L, Wang Y, Xiayu X, et al. Altered gut microbiota in a mouse model of 714 Alzheimer’s disease. J Alzheimer’s Dis. 2017;60(4):1241-1257. doi:10.3233/jad-715 170020 716 90. Niccolai E, Boem F, Russo E, Amedei A. The gut–brain axis in the 717 neuropsychological disease model of obesity: A classical movie revised by the 718 34 emerging director “Microbiome.” Nutrients. 2019;11(1):156. 719 doi:10.3390/nu11010156 720 91. Heath ALM, Haszard JJ, Galland BC, et al. Association between the faecal short-721 chain fatty acid propionate and infant sleep. Eur J Clin Nutr 2020 749. 722 2020;74(9):1362-1365. doi:10.1038/s41430-019-0556-0 723 92. Ba Z, Lee Y, Meng H, et al. Matrix Effects on the delivery efficacy of 724 Bifidobacterium animalis subsp. lactis BB-12 on fecal microbiota, gut transit time, 725 and short-chain fatty acids in healthy young adults. mSphere. 2021;6(4):e00084-726 21. doi:10.1128/msphere.00084-21 727 93. Lee Y, Ba Z, Roberts RF, et al. Effects of Bifidobacterium animalis subsp. lactis 728 BB-12® on the lipid/lipoprotein profile and short chain fatty acids in healthy 729 young adults: a randomized controlled trial. Nutr J. 2017;16(1):1-9. 730 doi:10.1186/s12937-017-0261-6 731 732 733 35 Table Legend 734 Table 1. PIO criteria 735 Table 2. Summary of trials selected for review 736 Figure Legend 737 Figure 1. Flowchart 738 Figure 1. Graph of the methodological quality of the trials included in the study. 739 Figure 2. Methodological quality of included clinical trials (n=25). 740 Figure 4. Forest plot showing the effects of GM modulation on the quality of sleep (n=18). 741 Figure 5. Forest plot reflecting the effects of GM modulation on sleep duration (n=5). 742 743 36 Table 1. PIO criteria 744 RCT: Randomized clinical trial 745 746 747 748 Parameter Criteria Population Adolescent and adult population (older than 18 years old) Intervention Probiotics and prebiotics formulations Outcomes Quality and duration of sleep Study design RCT Language Spanish or English 37 Table 2. Summary of trials selected for review 749 REFERENCE N INTERVENTION USED INTERVENTION DURATION (WEEKS) VARIABLES ANALYZED DATA COLLECTION TOOLS MAIN RESULTS PROBIOTIC Rode et al. (2022)57 Healthy subjects aged 20-28 (n=22) Lactobacillus helveticus R0052 (2×109 CFU), Lactiplantibacillus plantarum R1012 (8 × 108 CFU) and Bifidobacterium longum R0175 (7×107 CFU) Powder sachet 4 Anxiety Stress Depression Sleep quality Sleepiness Awakening HADS PSS STAI KSD There were no significant differences in sleep quality between the probiotic intervention and the placebo group Shafie et al (2022)58 Postmenopausal women aged 45-55 (n=66) Lactobacillus bulgaricus, Streptococcus thermophilus Bifidobacterium lactis and L. acidophilus Yogurt 6 Anxiety Stress Depression Sleep quality PSQI DASS-21 The use of probiotics had no effect on sleep quality Davoodabadi et al (2021)59 Females aged 18-40 suffered from cyclical mastalgia due to breast FCC (n=45) Lactobacillus acidophilus, L. fermentum, L. reuteri, and Bifidiobacterium bifidum 2x109 CFU Capsule 12 Depression Anxiety Sleep quality BDI BAI PSQI The probiotic group did not improve sleep quality West et al (2021)60 Night shift workers (n=87) Lactobacillus acidophilus DDS-1 and Bifidobacterium animalis spp. lactis UABla-12. 1x1010 CFU Capsule 2 Sleep quality PSQI Fitbit sleep time (min) Significant changes in sleep quality in the experimental groups Calgaro et al. (2021)61 Healthy subjects (n=33) Limosil actobacillus fermentum LF16 (DSM 26956), Lacticaseibacillus rhamnosus LR06 (DSM 21981), Lactiplantibacillus plantarum LP01 (LMG P-21021), and Bifidobacterim longum 04 (DSM 23233). 4×109 CFU/AFU. Sachet 6 Sleep quality PSQI Significant improvement in sleep quality Lee et al (2021)62 Healthy adults aged 19-65 (n=156) Lactobacillus reuteri NK33 and Bifidobacterim adolescentis NK98 2.5 × 109 CFU/500 mg Capsule 8 Sleep quality Symptoms of insomnia Mood BAI BDI-2 PSQI ISI SRI Significant reductions in sleep quality and severity of insomnia in the experimental group 38 Harnett et al (2021)42 Elite athletes (n=38) Ultrabiotic 60TMi and SBFloractivTMii Tablet 17 Sleeping hours Sleep quality Self-reported sleep quality Self-reported number of hours of sleep Sleep quality and quantity improved in the experimental group Moloney et al (2021)43 Students aged 18-40 (n=20) Bifidobacterium longum AH1714. 1x109 CFU Capsule 8 Depression measurements Sleep quality Mood PSS HADS-A, HADS-D PSQI The probiotic Bifidobacterium longum improved sleep quality and duration in the experimental group Wang et al (2021)44 Older adults (n=120) Bifidobacterium longum Lactobacillus acidophilus Enterococcus faecalis More of 1.0x107 CFU/g. Capsule 1 Postoperative cognitive impairment Postoperative sleep quality MMSE NRS No significant differences were found in sleep quality between the experimental and control group Marotta et al (2019)45 Healthy students (n=38) Lactobacillus fermentum LF16 (DSM 26956) L. rhamnosus LR06 (DSM 21981) L. plantarum LP01 (LMG P-21021) Bifidobacterium longum BL04 (DSM 23233) 4x109 CFU Powder 9 Presence and severity of anxiety symptoms and propensity to be anxious Occurrence and severity of depressive symptoms Mood Sleep quality LEIDS-R STAI BDI-2 POMS PSQI An increase in sleep quality was observed in the experimental group Wälivaara et al (2019)46 Healthy patients (n=61) Lactobacillus reuteri (DSM 17938 and ATCC PTA 5289). 2x108 live bacteria from the combined strains Tablet 2 Sleep Self-reported data on disturbed sleep The experimental group was associated with fewer nights of interrupted sleep Nishida et al (2019)27 Healthy students (n=60) Lactobacillus gasseri CP2305. 1x1010 bacterial cells per 2 tablets Tablet 24 Mental and physical condition Sleep quality STAI GHQ-28 HADS PSQI Intake of the probiotic significantly reduced sleep disorders Lei et al (2018)47 Patients with simple rib fracture (n=283) Lactobacillus casei Shirota. 6x109 CFU Milk 4 Sleep quality NRS There was no significant improvement in sleep quality Nishida et al (2017)63 Students aged 18-34 (n=32) Lactobacillus gasseri CP2305. 1x1010 CFU Canned milk 5 Mental and physical condition Sleep measurement and assessment GHQ-28 Zung-SDS HADS STAI PSQI Sleep quality improved in the probiotic group, shortened sleep latency, and increased sleep duration 39 Sawada et al (2017)49 Students (n.d.) Lactobacillus gasseri CP2305. 1.0x1010 CFU Powder 4 Mental health Sleep quality and disturbances GHQ-28 Zungs-SDS HADS STAI PSQI Probiotic ingestion prevented the development of sleep disorders and improved PSQI scores Kelly et al (2017)50 Healthy participants aged 20-33 years (n=29) Lactobacillus rhamnosus (JB-1) 1x109 CFU Capsule 8 Psychobiological responseiii to an acute stressor Cognitive performance Sleep quality CANTAB SECPT BDI-2 BAI PSS STAI PSQI Probiotic treatment did not significantly affect sleep quality relative to placebo treatment Takada et al (2017)26 Healthy students (n=124) Lactobacillus casei strain Shirota YIT 9029. 1.0x109 CFU/ml Milk. 11 Anxiety levels Sleep quality Sleep efficiency Healthy habits Psychological well-being STAI PSQI GHQ OSA HPI Positive effect on sleep duration and latency Culpepper et al (2016)51 Undergraduate students (n=581) Lactobacillus helveticus R0052 (LH), Bifidobacterium bifidum R0071 (BB) and B. longum spp. infantis R0033 (BI). 3×109 CFU. Capsule 6 Stress level Sleeping hours Self-reported stress scores Self-reported stress and levels of sleep Negative correlation between hours of sleep and self-reported stress values Wong et al (2014)52 IBS patients (n=42) Bifidobacterium longum, B. infantis and B. breve, Lactobacillus acidophilus, L. casei, L. delbrueckii ssp. bulgaricus and L. plantarum and Streptococcus salivarius ssp. thermophilus (VSL#3) 112.5 billion viable lyophilized bacteria Capsule 6 Sleep quality Daytime sleepiness HADS-A, HADS-D PSQI ESS No significant differences were found between the probiotic group and placebo Diop et al (2008)28 Volunteers with stress symptoms aged 18-60 (n=75) Lactobacillus acidophilus Rosell52 and Bifidobacterium longum Rosell175. (3 × 109 CFU) Sachet stick 3 Sleep disorders Self-reported levels of sleep Probiotic intervention did not change stressinduced sleep problems PREBIOTIC Saleh-Ghadimi et al (2022)53 Female obese type 2 diabetic mellitus patients aged 30–65 (n=63) Nutriose®06 (resistant dextrin) 8 Sleep quality PSQI Prebiotic intervention resulted in significant improvements in sleep quality when 40 compared to the placebo group Schaafsma et al (2021)54 Adults with sleep disturbances aged 30-50 (n=70) Product containing protein, GOS, vitamins and minerals Sachet 3 Sleep quality PSQI Treatment did not improve sleep quality Buigues et al (2016)29 Institutionalised older population without dementia (n=50) Darmocare Pre® (Inulin 3375mg and fructooligosaccharides 3488mg (FOS)) Powder 13 Effect of the prebiotic on the frailty syndrome MMSE AIS No significant changes in sleep quality were observed SYNBIOTIC Valle et al (2021)55 Military undergoing field training (n=65) Lactobacillus acidophilus LA-5 (2·1 × 108 CFU/g) Bifidobacterium animalis BB-12 (2·7 × 109 CFU/g) and inulin (23g/60g) Ice cream 4 Sleep quality Sleepiness WAM PSQI Sleep quality did not improve significantly in the experimental group Sleepiness decreased in the synbiotic-treated group Quero et al (2021)56 Professional soccer players and sedentary students (n=27) Gasteel Plus® (Bifidobacterium lactis CBP001010, Lactobacillus rhamnosus CNCM I-4036, Bifidobacterium longum ES1 (≥1×109 CFU) and FOS (200 mg)) Stick 4 Sleep quality Anxiety Stress Depression Fatigue HLPCQ. STAI PSS BDI BFI The intervention only improved the sleep quality of the athletes CFU: Colony-Forming Units; AIS: Athens Insomnia Scale; BDI-2: Beck Depression 750 Inventory; CANTAB: The Cambridge Neuropsychological Test Automated Battery; 751 GHQ: General Health Questionnaire; GHQ-28: The 28-item General Health 752 Questionnaire; WAM: Well-being, Activity, Mood questionnaire; BAI: The Beck Anxiety 753 Inventory; SRI: Stress Response Inventory; HADS-A, HADS-D: The Hospital Anxiety 754 and Disease Scale; LEIDS-R: Leiden Index of Depression Sensitivity-Revised test; ISI: 755 Insomnia Severity Index; MMSE: Mini-Mental State Examination; NRS: Numeric Rating 756 Scale; OSA: The Oguri-Shirakawa-Azumi sleep inventory; POMS: Profile of Mood State; 757 PSQI: Pittsburgh Sleep Quality Index; PSS: The Perceived Stress Scale; SECPT: The 758 Socially Evaluated Cold Pressor Test; STAI: The State Trait Anxiety Inventory; Zung-759 SDS: The Zung Self-Rating Depression Scale; HPI: Health Practice Index; ESS: Epworth 760 Sleepiness Scale; BFI: The Brief Fatigue Inventory; HLPCQ; The Sleep Quality in 761 Healthy Lifestyle and Personal Control Questionnaire; DASS-21: Depression, Anxiety, 762 and Stress Scale-21 items; IBS: irritable bowel syndrome; AFU: Active Fluorescent Units; 763 KSD: Karolinska Sleep Diary. GOS: galacto-oligosaccharides; FCC: fibrocytic change. 764 MADRS: Montgomery-Åsberg Depression Rating Scale; QIDS-SR16: Quick Inventory of 765 Depressive Symptomatology; GAD-7: Generalized Anxiety Disorder 7-item scale; MDD: 766 Major Depressive Disorder. 767 41 i This formula includes Lactobacillus acidophilus (CUL 21+CUL 60) 20 trillion CFU, L. rhamnosus (CUL 63) 15.55 trillion CFU, L. casei (CUL 06) 9.45 trillion CFU, L. plantarum (CUL 66) 3.15 trillion CFU, L. fermentum (CUL 67) 1.35 trillion CFU, Bifidobacterium animalis ssp. lactis (CUL 34) 6.55 trillion CFU, B. breve (CUL 74) 1.35 trillion CFU, B. bifidum (CUL 20) 345 million CFU, Streptococcus thermophilus (CUL 68) 2.25 trillion CFU. ii This formula consists of Saccharomyces cerevisiae (boulardii) 5 trillion CFU. iii This formula consists of 112.5 billion viable lyophilized bacteria: Bifidobacterium longum, B. infantis and B. breve, Lactobacillus acidophilus, L. casei, L. delbrueckii ssp. bulgaricus and L. plantarum and Streptococcus salivarius ssp. thermophilus.