Nutritional interventions in treating menopause-related sleep disturbances: a systematic review

Abstract Context Sleep disturbances are a core symptom of menopause, which refers to the permanent cessation of menstrual periods. Nutritional interventions may alleviate menopause-related sleep disturbances, as studies have shown that certain interventions (eg, tart cherry juice, or tryptophan-rich foods) can improve relevant aspects of sleep. Objective The aim of this systematic review was to examine the effect of nutritional interventions for menopause-related sleep disturbances, in order to inform the subsequent development of specific interventional trials and assess their potential as a treatment for menopause-related sleep disturbances. Data Sources Published studies in English were located by searching PubMed and PsycArticles databases (until September 15, 2022). Data Extraction Following full-text review, a final total of 59 articles were included. The search protocol was performed in accordance with PRISMA guidelines. Data Analysis A total of 37 studies reported that a nutritional intervention improved some aspect of sleep, and 22 studies observed no benefit. Most (n = 24) studies recruited postmenopausal women, 18 recruited menopausal women, 3 recruited perimenopausal women, and 14 recruited women from multiple groups. The majority of the studies were of low methodological quality. Due to the heterogeneity of the studies, a narrative synthesis without meta-analysis is reported. Conclusion Despite the large heterogeneity in the studies and choice of intervention, the majority of the identified studies reported that a nutritional intervention did benefit sleep, and that it is mainly subjective sleep that is improved. More high-quality, adequately powered, randomized controlled trials of the identified nutritional interventions are necessary. Systematic Review Registration PROSPERO registration no. CRD42021262367.


INTRODUCTION
Sleep has a major impact upon a number of health outcomes.Disrupted sleep, or sleep loss, can contribute to a range of deleterious health outcomes, including mortality, obesity, diabetes, and cardiovascular disease. 1,2For this reason, obtaining sufficient high-quality sleep is necessary for maintaining good physical and psychological health. 2 Although changes to subjective and objective sleep are commonly observed as a function of the normal aging process, 3 one specific life event that appears to have a direct impact upon sleep is the transition to menopause in women. 4Menopause is a complex physiological process and refers to the permanent cessation of menstrual periods due to ovarian follicular depletion, alongside changing hormonal levels of estrogen and progesterone, which can occur naturally or due to surgery, chemotherapy, or radiation. 5This typically occurs at approximately 50 years of age for Western women, although the timing can be influenced by lifestyle, race, and ethnicity. 6leep disturbances are considered to be a core symptom of menopause. 4Specific sleep disturbances which are commonly observed during menopause typically include subjective difficulties in falling asleep, awakening too early, excessive daytime sleepiness, and the clinical sleep problem of insomnia disorder (where individuals have difficulties in falling asleep, maintaining sleep, and awakening early, alongside the associated negative daytime consequences of these disturbances). 7,8Clinically, insomnia disorder is more common in women than in men, at a ratio of approximately 2:1, 9 and the prevalence of insomnia, which is the most commonly observed sleep disorder associated with menopause, increases around the time of menopausal onset. 10Also of relevance is the fact that menopause may result in alterations to circadian rhythmicity. 11ircadian rhythms refer to the oscillatory rhythms of approximately 24 hours that are displayed by various bodily physiological and behavioral processes (including hormones) and have a direct impact upon sleep timing and quality. 12,13It is well established that aging can affect circadian rhythmicity.For instance, aging affects chronotype, which refers to the preference for the timing of sleep and other daily activities 12 : older adults are more likely to prefer an earlier bedtime, and an earlier morning rise time, compared to young adults. 12dditionally, older adults appear to be more sensitive to the effects of a mismatch between an individual's desired and actual sleep timing (ie, when an individual is awake when their circadian rhythm favors sleep), compared to younger adults 14 ; this can impair sleep duration and sleep quality.Aside from the circadian changes that are observed as a function of normal aging, some evidence indicates that, after the menopausal transition, there is a shift toward "morningness," referring to an individual preference for earlier sleep timing and rise time (relative to "eveningness") 11 ; speculatively, this may be due to changes in hormonal secretion that occur as part of the menopausal transition. 11leep is a complex physiological process that is typically assessed using various subjective and/or objective measurement methods: subjective methods of sleep measurement can include, for example, questionnaires, estimates of habitual sleep duration, or sleep diaries (sleep logs). 3Objective methods of sleep measurement can include actigraphy, which relies on wrist-worn accelerometers to infer sleep and wake patterns based on movement, or polysomnography (PSG). 15PSG is the most accurate method of sleep measurement, as this method simultaneously assesses multiple physiological parameters, including overnight brain activity, in order to classify sleep into distinct stages. 16The concept of sleepiness is also relevant: sleepiness occurs when the brain is forced to transition from a state of arousal to a state of sleep. 17Sleepiness is primarily driven by increased sleep pressure caused by extended wakefulness or sleep deprivation, as well as by circadian rhythmicity. 13,17Sleepiness can also be measured subjectively, typically using self-report estimates or questionnaires, or objectively, where PSG is used to quantify sleepiness in the context of a nap opportunity, using the Multiple Sleep Latency Test (MSLT). 16t is likely that nutritional interventions and supplements may help to alleviate menopause-related sleep disturbances, since these interventions have been shown to improve relevant aspects of sleep.For instance, relative to placebo, tart cherry juice has been shown to improve objective sleep quantity and quality 18 ; tryptophan-rich foods have also been shown to improve subjective and objective sleep quantity and quality. 19Therefore, nutritional interventions and supplements are likely to represent one route by which menopause-related sleep disturbances can be treated, or alternatively, used to improve sleep.While a recently published narrative review article has concluded that nutritional interventions may improve menopause-related sleep disturbances, 20 the main purpose of that previous review was to provide a practical guide for the treatment of menopause-related sleep disturbances through nutritional changes.However, in order to inform the development of nutrition intervention studies with the specific intention of treating sleep disturbances associated with menopause, a systematic search of the literature is necessary, giving consideration to the potential mechanisms by which nutritional interventions can improve specific relevant sleep outcome variables.Therefore, the aim of the present review was to systematically examine the effect of nutritional interventions for menopause-related sleep disturbances, to inform the subsequent development of specific interventional trials.

METHOD
The search protocol was preregistered in PROSPERO (CRD42021262367) and performed in accordance with PRISMA guidelines. 21
Eligibility criteria.In the present review, nutritional interventions were defined as specific and measurable changes to diet, with the aim of affecting a relevant outcome measure.These included either specific dietary changes, or the use of nutritional or dietary supplements (in pill, tablet, powder, or liquid form).
Studies were eligible if they specifically assessed the effect of a nutritional intervention, or interventions, upon subjective and objective sleep, where sleep was assessed as either a primary or secondary outcome measure.This included studies where sleep was examined using subjective questionnaire measures, or clinically relevant questionnaire measures.This also included studies where subjective sleep continuity was measured using sleep diaries, from which standard measures of sleep continuity could be derived (eg, subjective sleep efficiency [SE%], number of awakenings [NWAK], wake after sleep onset [WASO], total sleep time [TST], and/or sleep onset latency [SOL]).In addition, studies that incorporated objective measures of sleep continuity (eg, using actigraphy or PSG), and studies that assessed the effects of an intervention upon objective sleep architecture (PSG) were included.Studies that assessed subjective or objective sleepiness (measured using questionnaires or objective methods, such as an MSLT) were also included.The studies were eligible if they included females who were perimenopausal, menopausal, or postmenopausal (as defined by the relevant research study).Unpublished studies and pre-print articles were not specifically sought, but were considered for inclusion if they were relevant, or if they were referenced in eligible studies.The PICOS criteria for the inclusion of studies are listed in Table 1.
Identified articles were excluded if (a) the effect of a nutritional intervention upon sleep could not be evaluated; or if they were: (b) duplicates; (c) review articles; (d) written in a non-English language; (e) conference abstracts; (f) opinion-based letters, or (g) animal studies.
Data extraction.Eligible articles were exported to EndNote X9.3 (Clarivate, London, UK), and the abstracts were independently screened against the inclusion criteria by D.P. and G.J.E.using a data extraction form.The methodological quality of studies, in relation to their main aims, was evaluated using the Mixed Methods Appraisal Tool (MMAT). 22,23Each study was assigned a rating from 0 (indicative of a very poor level of methodological quality) to 5 (the highest methodological quality) on the basis of the MMAT.The first 10% of identified papers and the MMAT evaluation was checked by another member of the study team (P.A.-M.).
Narrative synthesis.Due to the heterogeneity of studies and methodological approaches, it was not possible to conduct a meta-analysis.Therefore, a narrative description of the identified studies is provided, in line with Synthesis without Meta-analysis (SWiM) 24 reporting guidelines.

RESULTS
A total of 2134 potentially relevant articles were identified and screened.No additional results were identified at this point.Following screening, a total of 72 articles were chosen for full-text review.A final total of 59 articles were included (Figure 1).

Characteristics of studies
There was a large amount of heterogeneity in terms of the choice of nutritional intervention, which is summarized in Table 2.  The entified studies also displayed a large amount of heterogeneity regarding the study design, and the results of each study are shown in Table 3.  The mple sizes of the identified studies ranged from 18 to 2016 participants and were conducted between 2002 and 2022.The quality of the identified studies ranged from 1 to 5, and the most common MMAT score was 2, which is indicative of a low methodological quality.57,59,61,62,64,[69][70][71][72][73]75,78,79,83 Fourteen studies recruited women from multiple groups.26,30,33,37,39,40,45,49,54,60,67,74,80,81 Only 6 studies specifically assessed the impact of interventions upon women with self-reported sleep disturbances, 33,38 , 46- 48,75 2 of which involved a retrospective analysis of clinical records. 33,47 The majority of studies assessed the effect of the in32,35,38,42,43 , 46- 49,51,54,55,57,64,68-71,75,77,81,83 Additionally, 6 other studies reported the subjective impact upon sleep, but established questionnaire measures were not used.25,27,36,44,45,50 For instance, 2 studies used the selfreported presence and severity of insomnia 25,36 ; 1 reported the frequency of self-reported sleep disturbances (eg, how often nocturnal awakenings occurred in a previous time frame) 27 ; or a custom score that assessed the frequency of perceived nights of "bad sleep", as reported in 50 ; One study used an unspecified insomnia disorder measurement tool 44 ; and 1 study used an unspecified measure of subjective sleep quality and satisfaction.45 Only 4 studies concentrated on the impact of any intervention upon subjective sleep continuity measures.32,33,41,68 These included the average number of hours slept per month (derived from a daily diary), 68 the subjective number of awakenings (also derived from a diary), 32 and, in one study, where a sleep disturbance score was calculated, alongside subjective sleep quality (in terms of sleep duration, the ease of sleep onset, the number of awakenings per night, and perceived satisfaction with sleep).33 A further study used an unspecified self-report measure that assessed total sleep time, wake after sleep onset, and other non-standard sleep continuity parameters (such as "frequency of wake after sleep onset" and "longest sleep time").79 Finally, only 3 studies measured the effect of any intervention upon objective sleep: 1 study used actigraphy as the main outcome 60 and 2 studies used 1 or 2 nights of PSG, respectively. 53,64 Oveall, a total of 37 studies reported that the nutritional intervention improved some aspect of sleep, 25,28- 30,32-36,38-41,43,44,46-48,50,51,53,54,56-58,61,63,64,67,69,72-78 and a total of 22 studies observed no benefit upon sleep, or sleepiness.26,27,31,37,42,45,49,52,55,59,60,62,65,66,68,70,71,[79][80][81][82][83] Of the 37 studies that showed a positive effect upon sleep, this included a total of 18 studies where improvements were observed using a menopause symptom questionnaire, such as the Greene Climacteric Scale (GCS), Menopause Rating Scale (MRS), or Kupperman Index (KI), which assessed some element of sleep such as the presence or absence of sleep disorders.28,29,34,36,39,40,43,58,61,63,65,67,[72][73][74]76,78,82 These 37 studies also included a total of 10 studies in which a benefit upon subjective sleep quality was observed, as measured using a validated and commonly used measure of sleep quality (PSQI).32,35,38,[46][47][48][49]51,75,78 Four studies found that nutritional interventions reduced the severity of the sleep disorder insomnia, 54,57,69,77 as measured using established questionnaires including the Athens Insomnia Scale (AIS) 54,77 and the standard or modified ISI.57,69 Only 1 study reported an improvement in subjective sleep continuity, which was measured using a 1-day sleep diary. 32 A otal of 6 studies found improvements when subjective sleep was assessed using unspecified, or custom, measures of assessment.25,30,33,41,44,50 This included studies using selfreport measures of insomnia 25,44 or self-report            assessment of the frequency of nights of bad sleep during the previous night 50 , or subjective sleep quality.30 Other studies have found a benefit upon unspecified measures of sleep quality and quantity, 33 or subjective measurement of sleep continuity which has included standard continuity variables such as total sleep time or wake after sleep onset (WASO), in addition to custom variables including "nighttime sleep time", "frequency of WASO", "longest sleep time" and "sleep onset time".41 Finally, only 1 study specifically examined the impact upon objective sleep, which was measured using PSG.In this study, the intervention (black cohosh) resulted in lower WASO and higher SE%, which represents a reduction in nocturnal wake duration and increased sleep quality respectively, relative to placebo; however, no corresponding benefit to subjective sleep quality (PSQI) was observed. 6467,74 The studies are summarized in more detail in Table 3.

DISCUSSION
Despite the heterogeneity of the identified studies, and in the choice of intervention (Table 2), the majority of the studies reported that a nutritional intervention did benefit sleep.While these results primarily indicate that it is subjective sleep that is improved by nutritional interventions, interestingly, 2 studies are also suggestive of improvements to objectively measured sleep, which was assessed using PSG. 53,64The limitations of the identified studies, the implications of the studies that have observed positive effects, and potential directions for future nutritional trials aimed at alleviating or preventing menopausal-related sleep disturbances are summarized below.
Isoflavones, either used as a standalone intervention, or in combination with other agents, appear to be particularly promising, as benefits have been found upon subjective sleep. 25,29,36,50,53,54,56,574][55] The dosage may influence the degree of the therapeutic effect, as 1 study has found that a higher dose (25 mg daily) was more effective than a lower dose (12.5 mg) in the treatment of self-reported insomnia symptoms. 54In terms of other interventions, 3 studies indicated that soy or soybean-based interventions were beneficial in terms of subjective sleep 35,58,61 ; the effects of soy are typically attributed to their high isoflavone content. 84dditionally, black cohosh may warrant further investigation: in 1 study it was found to improve subjective sleep problems (assessed on the basis of a menopausal symptom assessment), 63 and in another study, it improved objective sleep, in terms of PSG-assessed markers of nocturnal wake duration (WASO) and objective sleep quality (SE%), although not subjective sleep quality, relative to placebo, at six months followup. 64lthough other interventions, including pomegranate seed oil, 73 pollen extracts, 51 organic Maca, 76 jujube seed capsule, 75 and salvia extract, 78 appear to benefit subjective sleep, and high doses of proanthocyanin may benefit subjective insomnia symptoms, 77 in all cases, only 1 study has assessed the impact of each of these interventions.A variety of other studies have indicated that combined nutritional products benefit subjective self-reported insomnia, sleep quality, sleep disturbances, and nocturnal awakenings, [28][29][30][32][33][34]36,74 and similarly, the use of melatonin, combined with a reduced caloric intake, reduced subjective insomnia severity. 69 Finally,only 5 studies have specifically focused on the use of nutritional interventions in women with clinically significant poor sleep quality, and insomnia, respectively.38,[46][47][48]53 In the first study, valerian and lemon balm improved subjective sleep quality in women with poor sleep quality, 38 which was defined as PSQI scores of !5, which is an accepted cut-off value for poor sleep.However, in this study, the duration of the valerian and lemon balm usage was not stated.38 Similarly, in another study, jujube seed capsules improved subjective sleep quality (expressed as reduced PSQI scores) at a 21-day follow-up time point in women with baseline PSQI scores of !5. 75 Finally, in women with PSQI scores of >6, a combination of 4 traditional Chinese medications (Guizhi Gancao Longgu Muli Tang) improved sleep quality at follow-up.48 In postmenopausal women with self-reported insomnia, isoflavones benefited sleep, 53 and another study found that a combination of herbal ingredients reduced PSQI scores at follow-up, although this was compared to an unspecified control group, and details regarding the PSQI scores were not stated.47 However, in both studies, the definition of insomnia was not clearly stated; consequently, it is not clear whether women with insomnia symptoms were recruited, or if participants had an appropriate clinical diagnosis of insomnia disorder.
Overall, despite the heterogeneity in the interventions, and the relatively low number of studies, the most consistent finding is that isoflavone-based interventions do appear to benefit subjective sleep, 25,29,36,50,53,54,56,57 and, as mentioned, 1 study has indicated that there might be a dose-response effect upon sleep. 54That said, one limitation is that the precise mechanism by which isoflavones could affect sleep is currently unclear and will require further investigation. 85One speculative possibility is that, because isoflavones are phytoestrogens, which bind to estrogen receptors, and as isoflavones may reduce menopausal symptoms, 86 isoflavones may improve sleep by alleviating or improving menopausal symptom severity.This does appear to be a plausible explanation, since it is known that sleep disturbances are associated with menopausal symptoms. 7,8lack cohosh also appears to improve subjective and objective sleep [63][64][65] ; however, as with isoflavones, the exact mechanism of action for this intervention upon sleep is not well established.However, it is possible that black cohosh can affect the neurotransmitters that modulate sleep/wake regulation, including serotonin (5-HT) and c-aminobutyric acid (GABA). 64lthough only 3 studies have assessed the impact of this intervention, the results of all 3 were positive, and this intervention is certainly likely to be worthy of further investigation in the context of a larger-scale trial.Finally, as stated, various other specific interventions show promise in terms of benefitting subjective sleep 51,73,[75][76][77][78] ; however, to date, only 1 study of each of these interventions has shown that there is a benefit.Overall, larger, higher-quality studies need to be conducted to replicate these findings.

Nutritional interventions that do not benefit sleep
It should also be noted that a number of the identified studies have demonstrated that a range of nutritional interventions do not benefit sleep or sleepiness (Table 2).However, for some interventions, only 1 or 2 studies investigating the subsequent effect upon sleep have been conducted.For instance, this was the case in studies that investigated resveratrol and trans-resveratrol, 70,71 Gincosan, 79 Labisia pumila var.alata water extract, 80 pine bark extract, 82 probiotic yogurt, 83 and omega-3. 81As was the case with the nutritional interventions that did observe a positive effect upon sleep, there was a great deal of heterogeneity between studies in terms of the choice of intervention, the chosen outcome measure, and the menopause status of participants.It is likely that the primary reason for these null results is due to such a wide range of methodological limitations.Although these specific limitations will be discussed in detail in the following section, they have included problems such as lack of trial preregistration, poor methodological quality, and in the choice and suitability of the sleep outcome measure.
Despite the apparent heterogeneity in the negative studies, of the studies where no benefit was observed at all, there are 2 main similarities.Firstly, sleep or sleepiness was not considered to be the primary outcome measure, 70,71,[79][80][81][82] and, notably, in 3 studies the primary aim was to demonstrate the efficacy and safety, or efficacy and tolerability, of the intervention, [80][81][82] rather than to assess the clinical effectiveness upon a prespecified primary outcome.0][81][82] Overall, as the focus of these studies was to only assess the impact upon sleep as part of a menopausal symptom profile (as a secondary measure), and as there was no sound mechanistic justification for an expected benefit to sleep, it is perhaps unsurprising that no benefit was observed.As an additional point, the majority of the null studies recruited only postmenopausal women, 70,71,79,83 and included studies in which the intervention was Gincosan, 79 probiotic yogurt, 83 or resveratrol and trans-resveratrol. 70,71Despite the apparent lack of impact of the interventions, this does perhaps suggest that future work should consider if menopausal status is likely to have an impact upon the effectiveness of a given intervention.

Limitations of identified studies
Despite the fact that a diverse range of nutritional interventions do appear to benefit menopause-related sleep disturbances, the identified studies have a number of limitations that should be considered and addressed.
0]80 This was the case irrespective of whether positive or null effects were observed.44]54,57,67,78 Another potential problem is that many of the clinical trials that were identified in the present review did not appear to have been preregistered. 25- 27,30,31,34,36,41,44,52,53,56,59,64,68,72,75-77,79-82The lack of preregistration is problematic, as the preregistration of clinical trials is necessary to ensure transparency, and to also minimize the impact of, or prevent, publication bias and the selective reporting of results upon trial completion. 87In some studies, full details regarding the dosage, contents, or the frequency of usage of each intervention was not stated, 33,41,48,[65][66][67] which is likely to hinder replication studies.
Another problem is that the majority of the identified research studies used combined interventions, where the specific intervention or supplement was comprised of multiple potentially "active" ingredients. 25- 36,39-41,43,44,46,47,57A clear limitation of this approach is that, generally, a sound rationale for the specific choice of dosage and specific combination of the intervention was not provided when the results were reported; in some instances, these combined multiple "active" ingredients that can affect sleep in their own right.For instance, in 1 study, participants were given a morning menopause formula consisting of Panax ginseng, black cohosh, soy, and green tea extract, and an evening menopause formula consisting of black cohosh, soy, kava, hops, and valerian extracts. 32Previous work has demonstrated that l-theanine, which is an amino acid contained within tea, may enhance GABA levels; GABA has a key role in the regulation of sleep and wake 88,89 ; additionally, valerian is a sleep-promoting medication, potentially due to the lignan derivate olivil binding to adenosine A1 and GABA-A receptors, subsequently resulting in sedative effects. 90Consideration should be given as to whether or not such combinations of ingredients or compounds within the interventional product have the potential to cause additive or synergistic effects, since this is very likely to affect sleep.Finally, while 1 study reported that a proprietary intervention ("JuicePLUSþ") was found to be effective in terms of reducing subjective sleep problems at follow-up, there was no description of what was contained in the intervention in the study report 74 ; therefore, the underlying mechanism of action is not clear.When proprietary interventions are used, the contents of the intervention should be clearly stated, and the proposed mechanism of action should be fully explained and justified.
A further issue is that, when nonpharmacological interventions are used alongside nutritional interventions, this may result in additive effects, where the impact of the nutritional intervention alone is not wellunderstood.This was the case in 2 of the identified studies, in which it was not possible to isolate the effects of the nutritional intervention from the effect of the additional interventions that were used at the same time. 53,81In 1 study, omega-3 was administered alongside a placebo, but the participants were concurrently randomized to receive yoga, exercise, or usual physical activity 81 ; although there was no difference between the omega-3 and placebo groups at the 12 weeks follow-up, importantly, both groups showed improved sleep quality, and reduced insomnia severity. 81While mechanistically there is reason to support the administration of omega-3, due to its potential role within the melatoninproducing pineal gland, which is a key region involved in sleep/wake regulation, 91,92 meditative movement interventions such as yoga, and physical activity, can also both affect sleep, 93,94 and this may have masked the true effect of omega-3 upon sleep.
Similarly, another study investigated the effects of isoflavones in menopausal women with insomnia disorder 53 and although participants received a sleep education lecture at the beginning of the study, which summarized sleep hygiene, menopausal symptoms and general healthcare, sleep education has been shown to result in improvements to sleep in its own right. 95This study observed that whilst there was no difference in insomnia at the 2-month or 4-month follow-up time points, relative to a placebo condition, objective SE% improved in the isoflavones group, and importantly, both groups showed a significant reduction in the percentage of women reporting moderate or severe insomnia at 4 months follow-up. 53Given that it is possible that, for instance, both exercise and sleep hygiene can influence sleep, comparator conditions need to be chosen with caution.
One final issue is in regard to the choice of outcome measures.As stated, many studies focused upon sleep as a secondary outcome, 26,45,52,59,62,63,72,73,79,81 or as part of various menopausal symptoms. 29-31,39,41- 44,54,57,67,78However, an issue is that many studies assessed whether or not the intervention was effective simply by reporting the percentage of study participants who reported subjective improvements, or the absence of symptoms at a follow-up time point, and not by comparing subjective or objective numerical improvements using an appropriate statistical test. 29,50,58,71,74imilarly, other studies simply directly compared intervention and non-intervention groups at follow-up time points in order to judge the effectiveness of the interventions, without incorporating baseline assessments from both groups into the statistical tests 62,64 ; this is problematic and may contribute to the reason as to why many of the studies observed null results. 26,27,31,37,52,55,59,60,62,65,66,68,70,71,79,81,82To accurately determine whether or not a given intervention has a beneficial effect upon sleep, future studies should use appropriate statistical tests and clearly report accompanying measures of effect size.

Suggestions for future research
Overall, these results clearly indicate that high-quality, adequately powered, randomized controlled trials of nutritional interventions are necessary for research into the treatment of menopause-related sleep disturbances.Given that qualitatively, the majority of the studies that were identified in the present review were of a poor methodological quality, future trials need to have clear, well-defined, and appropriate sleep outcome measures.
In terms of specific future research directions, aside from the high-quality trials of the specific interventions identified in the previous section, perhaps one of most important is that ideally, nutritional interventions should have a sound underlying mechanistic justification in relation to the target symptom (eg, valerian may be sleep-promoting due to its impact upon adenosine A1 and GABA-A receptors 90 ).It is perhaps surprising that interventions which have been shown to improve sleep in other groups have not yet been trialled in relation to menopause-related sleep disturbances, such as tart cherry juice, 18 which does appear to improve objective sleep and promote exogenous melatonin levels, or tryptophan-rich foods. 19These could potentially be trialed in the short term.
Future studies should also consider that the pathophysiology of sleep disturbances (eg, the disruption to sleep) in relation to menopause is complex, and this should inform the choice of design and participant group.First, sleep disturbances can occur because of menopause, but they can also occur as a secondary cause: physiological or psychological changes that happen alongside menopause, such as vasomotor symptoms, depression, and/or anxiety, 7,96 trigger the sleep disturbances.For instance, the vasomotor symptom of hot flashes is associated with sleep disturbances in menopausal women. 10Although these effects may be difficult to disentangle due to the bidirectional associations between, for example, sleep, anxiety, and depression, 97- 100 well-designed studies which carefully consider the nature and time course of the sleep disturbance in participants (eg, primary vs secondary) may be able to more accurately determine if the chosen intervention is effective.Finally, future studies should consider methodological design choices such as follow-up duration, or the dosage, in order to assess and optimize both the short-term and long-term effectiveness of interventions.For instance, from the identified studies, it is likely that the dosage of the intervention may have an impact upon the subsequent effectiveness. 54uture trials should also use both subjective and objective measures of sleep as outcome measures.Although subjective sleep quality can be assessed using questionnaires such as the PSQI, 101 subjective sleep diaries can also be used to assess sleep; these involve individuals recording the times at which they sleep and wake up, sleep duration, and the frequency and duration of nocturnal awakenings, typically for a minimum period of 1 week. 102Particular advantages of sleep diaries are that they are cheap and easy to use, and that measures of sleep continuity can easily be derived from completed diary entries (eg, total sleep time, time in bed, and SE%; the latter is a marker of subjective sleep quality) 3 ; this is importantas insomnia disorder is diagnosed based on the subjective experience of sleep, and insomnia disorder is common in menopause. 7One potential advantage of sleep diaries is that they can indicate which particular aspect of sleep continuity is improved by the intervention; this would researchers to determine if, for example, an intervention reduces the frequency and duration of nocturnal awakenings, or if it improves SE%.Similarly, where studies intend to assess the impact of an intervention upon a specific sleep disorder (eg, insomnia disorder), care must be taken to ensure that participants are recruited in accordance with recognized diagnostic criteria, such as those contained within the Diagnostic and Statistical Manual of Mental Disorders, or the International Classification of Sleep Disorders. 103,104ctigraphy could also be employed as a marker of objective sleep, since this is generally considered to be the most suitable alternative to PSG when used for these purposes. 15Despite the relatively low cost and ease of use of this technique, it is perhaps surprising that only 1 study has used actigraphy as an outcome measure. 60A further advantage of actigraphy is that in addition to standard measures of objective sleep continuity, nonparametric methods of activity analysis can be used as outcome measures; this is important as nonparametric methods of actigraphy are more sensitive to change and may therefore be used as a more sensitive outcome measure in nutritional trials. 105[108] CONCLUSION Despite the fact that sleep disturbances are a core symptom of menopause, the results of the present systematic review indicate that the majority of the studies that have been conducted are of a low methodological quality and have issues including a lack of preregistration and potentially unsuitable statistical analysis.However, despite the large heterogeneity in the choice of intervention and in study designs, the majority of the identified studies showed that a nutritional intervention improved sleep, and specifically, that the interventions improved subjective sleep.The most promising nutritional interventions include isoflavones, soy or soybean-based interventions, and black cohosh.High-quality, adequately powered, randomized controlled trials of these interventions are now necessary, with appropriate outcome measures.These should be trialed alongside other current or future nutritional interventions or supplements that have a sound mechanistic justification for the improvement of subjective or objective sleep.

Table 1
PICOS criteria for inclusion of studies Intervention Specific dietary changes, or the use of nutritional or dietary supplements (in pill, tablet, powder, or liquid form) Comparator Clinical trials with an appropriate control group, or in the case of observational studies, an appropriate pre-intervention/post-intervention comparison Outcome Subjective and objective sleep, or subjective and objective sleepiness (assessed as either a primary or secondary outcome measure) Study design Observational, experimental, randomized controlled trial (of any design)

Table 2
Summary of interventions

Table 3
Summary of identified studies