Crocus Sativus for Insomnia: A Systematic Review and Meta-Analysis

Insomnia is a common complaint affecting human performance in daily life activities. This study aimed to analyze the effects of Crocus sativus on insomnia. Methods: PRISMA guidelines were used according to the PICOS model. A systematic search of PubMed/Medline and the Cochrane Library was undertaken for literature until December 2021. A random effects model was used with I2 statistic to assess heterogeneity and a GRADE assessment was used to assess the quality of the outcomes. Eight articles were included, involving 431 participants. Crocus sativus reduced insomnia severity (SMD: 0.53; 95%CI: −0.05 to 1.11; I2 statistic = 59%; p = 0.08) and increased sleep quality (SMD 0.89, 95% CI 0.10 to 1.68; I2 statistic = 90%; p = 0.03; 6 studies, 308 participants, very low-quality evidence) and duration (SMD: 0.57; 95%CI: 0.21 to 0.93; I2 statistic = 40%; p = 0.002; 5 studies; 220 participants, moderate-quality evidence) compared with the placebos. Although there is limited evidence of a very low- to moderate-quality, Crocus sativus may benefit people with insomnia. This non-pharmacological intervention may reduce the chance of adults with insomnia taking sedative–hypnotic medication, thus reducing dependency and withdrawal symptoms.


Introduction
Insomnia is one of the most common medical complaints. Insomnia may be experienced at all stages of adulthood, and the problem is chronic for millions. Furthermore, it can be a symptom of other disorders, such as depression or a primary disorder. Most often, insomnia is defined by the presence of an individual report of difficulty with sleep [1]. Insomnia that lasts for one year or more is a risk factor for the development of depression [2]. The presence of long sleep latency, frequent awakenings during nighttime, prolonged periods of wakefulness during the sleep period, or even frequent transient arousals is defined as evidence of insomnia [1].
The recent definition by the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) states that it makes no distinction between primary and comorbid insomnia. The DSM-5 defines insomnia as dissatisfaction with sleep quantity or quality, associated with one (or more) symptoms such as difficulty initiating sleep, difficulty maintaining sleep, which is characterized by frequent awakenings or problems returning to sleep after awakenings, and early-morning awakening with an inability to return to sleep. According to the other criteria, sleep disturbance also causes clinically significant distress or impairments in social, occupational, educational, academic, behavioral, or other important areas of functioning. The sleep difficulty occurs at least three nights per week, is present for at least three months, occurs despite adequate opportunity for sleep, cannot be explained by, and does not occur exclusively during, another sleep-wake disorder, is not attributable

Materials and Methods
Our protocol was registered in PROSPERO with the registration number CRD42022308316. The research was conducted according to the PRISMA guidelines. Relevant studies were identified by searching the following databases: Cochrane Central Register of Controlled Trials (CENTRAL) and MEDLINE (PubMed) for randomized control trials (RCTs) not later than 31 December 2021, which compare Crocus sativus for insomnia with placebos. The studies were reviewed using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines [16].

Literature Searching Strategies
We searched the Cochrane Central Register of Controlled Trials, CENTRAL (December 2021) and MEDLINE (2010 to December 2021). We used the search strategy in Appendix A to search MEDLINE and CENTRAL. We restricted the publications to the English language only. We checked the reference list of identified RCTs and review articles to find unpublished trials or trials not identified by electronic searches. We searched for ongoing trials through the World Health Organization (WHO) International Clinical Trials Reg-istry Platform (ICTRP), http://www.who.int/ictrp/en/ (accessed on 5 January 2022) and www.clinicaltrials.gov (accessed on 5 January 2022).

Inclusion and Exclusion Criteria
We included randomized control trials (RCTs) comparing Crocus sativus for insomnia with placebos in adults with insomnia. We included blinded and open-label studies. The interventions included Crocus sativus extract, either herbal oil or oral capsule. The comparisons included placebo, no treatment, or standard treatment. Insomnia must have been diagnosed by clinicians. The outcomes were divided into primary and secondary. The primary outcomes included insomnia severity and sleep quality, including subjective sleep quality, sleep latency, sleep duration, sleep efficiency, sleep disturbances, use of sleeping medications, and daytime dysfunction. The secondary outcomes included anxiety level, depression level, restorative sleep, and quality of life, such as physical and mental wellbeing.

Quality Assessment
Two review authors (M.P.M., M.N.N.) scanned the titles and abstracts from the searches and obtained full-text articles when they appeared to meet the eligibility criteria or when there was insufficient information to assess the eligibility. We independently evaluated the eligibility of the trials and documented the reasons for exclusion. We resolved any disagreements between the review authors by discussion. We contacted the authors if clarification was needed.
From each of the selected studies, we extracted: the study setting, participant characteristics (age, sex, ethnicity), methodology (number of participants randomized and analyzed, duration of follow-up), method for assessing insomnia, and occurrence of related adverse events. We assessed the risk of bias based on random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessors, completeness of outcome data, selectivity of outcome reporting, and other biases [17]. We resolved any disagreements by discussion. We assessed the quality of evidence for primary and secondary outcomes according to the GRADE methodology [18] for risk of bias, inconsistency, indirectness, imprecision, and publication bias, classified as very low, low, moderate, or high.

Statistical Analyses
Meta-analyses were performed using Review Manager 5.4 software (RevMan 2020). We used a random-effects model to pool data. Thresholds for the interpretation of the I 2 statistic can be misleading since the importance of inconsistency depends on several factors. We used the guide to interpret the heterogeneity as outlined: 0% to 40% might not be important; 30% to 60% may represent moderate heterogeneity; 50% to 90% may represent substantial heterogeneity; and 75% to 100% would be considerable heterogeneity [17].
We assessed the presence of heterogeneity in two steps. First, we assessed obvious heterogeneity at face value by comparing populations, settings, interventions, and outcomes. Second, we assessed statistical heterogeneity using the I 2 statistic [17].
We measured the treatment effect for dichotomous outcomes using risk ratios (RRs) and absolute risk reduction, and for continuous outcomes, we used mean differences (MDs), both with 95% confidence intervals (CIs). The planned subgroup analyses were routes of saffron administration and the presence of comorbidity.
We checked the included trials for unit of analysis errors. Unit of analysis errors can occur when trials randomize participants into intervention or control groups in clusters, but analyze the results using the total number of individual participants. We contacted the original trial authors to request missing or inadequately reported data. We performed analyses on the available data if missing data were not available. We performed a sensitivity analysis to investigate the impact of risk of bias for sequence generation and allocation concealment in the included studies. If there were sufficient studies, we intended to use funnel plots to assess the possibility of reporting biases, small study biases, or both.

Results of the Search
A total of 346 records were found from the comprehensive database search (Figure 1). Upon elimination of duplicates, a total of 46 records were screened for eligibility. We reviewed full copies of 22 studies; we identified 8 articles as possibly meeting the review inclusion criteria, and 5 that were not eligible for inclusion. Three were reviews [19][20][21], one trial did not include an outcome of interest [22], and in one trial the intervention was a part of the control [23]. Therefore, we included eight studies.
analyses on the available data if missing data were not available. We performed a sensitivity analysis to investigate the impact of risk of bias for sequence generation and allocation concealment in the included studies. If there were sufficient studies, we intended to use funnel plots to assess the possibility of reporting biases, small study biases, or both.
Insomnia severity and sleep quality were the primary outcomes of our review. The Insomnia Severity Index and Insomnia Symptom Questionnaire assessed the severity of the insomnia. The Insomnia Severity Index is a seven-item self-report questionnaire assessing the nature, severity, and impact of insomnia [32]. It assesses the perceived severity of difficulties initiating sleep, staying asleep, early morning awakening, satisfaction with current sleep patterns, and interference with daily functioning. A 5-point Likert scale is used to rate each item yielding a total score ranging from 0 to 28, with higher scores indicating more severe insomnia [33].
The Insomnia Symptom Questionnaire is a 13-item self-report instrument designed to identify insomnia. The items assessed the presence of a complaint of difficulty initiating or maintaining sleep, or a feeling that the sleep was nonrestorative or unrefreshing; the frequency of complaints and the duration of these symptoms; and the severity of daytime correlates of the sleep complaint. Higher scores indicate more severe insomnia [34].
Sleep quality in this review was assessed using the Pittsburgh Sleep Quality Index and the Pittsburgh Sleep Diary. The Pittsburgh Sleep Quality Index is a common tool that consists of a 19-item questionnaire used to measure sleep quality complaints [35][36][37]. It is divided into seven components that assess habitual sleep duration, nocturnal sleep disturbances, sleep latency, sleep quality, daytime dysfunction, sleep medication usage, and sleep efficiency. Every seven components are scored between 0 and 3. The total scores range from 0 to 21, with higher scores indicating poorer sleep quality [36,37].
The Pittsburgh Sleep Diary is a 14-item sleep diary composed of total sleep time, sleep latency, the number of awakenings after sleep onset, sleep quality, mood, and alertness. It consists of a 5-point Likert rating ranging from very bad to very good, with higher scores indicating better sleep quality [38].
Secondary outcomes included in our review were anxiety level, depression level, restorative sleep, and quality of life. Anxiety level was reported in two trials [26,28]. Anxiety level was measured using the Profile of Mood States in one trial [26] and one trial used the Spielberg Anxiety Inventory [28].
The Profile of Mood States is a widely applied measure for assessing an individual's mood. It consists of 35 items with 4 scales: dejection/anxiety (14 items), fatigue (7 items), vigor (7 items), and anger (7 items). Determining anxiety level using the POMS is a way to assess a patient's current mood state, which is rated on a 4-point scale (from not at all to extremely). The higher the score, the higher the level of anxiety [39,40]. The DASS-21 is a validated self-reported measure assessing stress, anxiety, and depression symptoms. It is composed of 21 questions which are rated on a 4-point scale. Subscale scores for depression, anxiety, and stress are calculated. The higher the scores, the worse the anxiety. For depression levels, the research tools used were Profile of Mood States, Depression, Anxiety, Stress-Scale 21, and Beck Depression Inventory-II. The higher the scores, the worse the depression [41].
The other secondary outcomes, including restorative sleep, were measured by a Restorative Sleep Questionnaire (RSQ). It was evaluated by one study that reported two results [25]. The RSQ is a validated 11-item questionnaire with good psychometrics that assess the refreshing quality of sleep by asking participants to rate it on a 5-point scale. It measures feelings of tiredness, mood, and energy. The RSQ could distinguish between healthy controls, patients with primary insomnia, and insomnia patients with isolated non-restorative sleep complaints. Higher scores indicate better restorative sleep [42].
For assessing the quality of life, the research tool used in this review was the 36-Item Short Form Survey (SF-36) [43]. The effect of Crocus sativus on SF-36 results was evaluated by a single study [27]. The SF-36 is an instrument that assesses health-related quality of life, consisting of two main domains: physical and mental. The physical domain is represented by the Physical Component Summary (PCS), and the mental domain is represented by the Mental Component Summary (MCS). The SF-36 measures eight scales: physical functioning, physical limitation, bodily pain, general health, vitality, social functioning, emotional limitations, and mental health. All these scales contribute to the scoring of two domains, physical and mental. High scores define a more favorable health state [43].
The assessment of the risk of bias is shown in Figures 2 and 3. Figure 2 shows the proportion of studies assessed as low, high, or unclear risk of bias for each risk of bias indicator. Figure 3 shows the risk of bias indicators for individual studies. The assessment of the risk of bias is shown in Figures 2 and 3. Figure 2 shows the proportion of studies assessed as low, high, or unclear risk of bias for each risk of bias indicator. Figure 3 shows the risk of bias indicators for individual studies.  For allocation concealment, the selection bias in five studies was unclear risk [25,26,[29][30][31], and all used computer-generated randomization. Three studies had unclear risks [24,27,28]. Allocation concealment methods were described in two studies [25,26] as  The assessment of the risk of bias is shown in Figures 2 and 3. Figure 2 shows the proportion of studies assessed as low, high, or unclear risk of bias for each risk of bias indicator. Figure 3 shows the risk of bias indicators for individual studies.  For allocation concealment, the selection bias in five studies was unclear risk [25,26,[29][30][31], and all used computer-generated randomization. Three studies had unclear risks [24,27,28]. Allocation concealment methods were described in two studies [25,26] as For allocation concealment, the selection bias in five studies was unclear risk [25,26,[29][30][31], and all used computer-generated randomization. Three studies had unclear risks [24,27,28]. Allocation concealment methods were described in two studies [25,26] as low risk. However, six studies did not mention allocation concealment [24,[27][28][29][30][31]. All eight studies included in this review used a placebo control. Four studies described the two components, blinding of participants and blinding of outcome assessment [25,27,29,31], as low risk. In four studies, the blinding of personnel was unclear [24,26,28,30]. The blinding of outcome assessments was not described in four studies [24,27,29,31] and was categorized as unclear risk. We judged blinding of personnel and outcome assessment as an unclear risk of bias. For incomplete data, six studies were assessed as low risk for attrition bias [24,25,27,[29][30][31]. The incomplete data outcome was not described in the two studies [26,28]. Thus, we judged them as presenting an unclear risk of bias. All eight studies reported the outcomes specified in their methods section [24][25][26][27][28][29][30][31] for selective reporting and were assessed as low risk for selective reporting. We detected no other potential sources of bias.

Insomnia Severity
There was a small reduction in insomnia severity in the Crocus sativus groups compared with the placebo groups (SMD: 0.53; 95%CI: −0.05 to 1.11; I 2 statistic = 59%; p = 0.08; 2 studies, 118 participants, moderate-quality evidence) [25,29] (Figure 4, Table 2). low risk. However, six studies did not mention allocation concealment [24,[27][28][29][30][31]. All eight studies included in this review used a placebo control. Four studies described the two components, blinding of participants and blinding of outcome assessment [25,27,29,31], as low risk. In four studies, the blinding of personnel was unclear [24,26,28,30]. The blinding of outcome assessments was not described in four studies [24,27,29,31] and was categorized as unclear risk. We judged blinding of personnel and outcome assessment as an unclear risk of bias. For incomplete data, six studies were assessed as low risk for attrition bias [24,25,27,[29][30][31]. The incomplete data outcome was not described in the two studies [26,28]. Thus, we judged them as presenting an unclear risk of bias. All eight studies reported the outcomes specified in their methods section [24][25][26][27][28][29][30][31] for selective reporting and were assessed as low risk for selective reporting. We detected no other potential sources of bias.

Insomnia Severity
There was a small reduction in insomnia severity in the Crocus sativus groups compared with the placebo groups (SMD: 0.53; 95%CI: −0.05 to 1.11; I 2 statistic = 59%; p = 0.08; 2 studies, 118 participants, moderate-quality evidence) [25,29] (Figure 4, Table 2).  [25,29]. Note: Line indicates confidence interval and the green square indicates effect estimate. Black diamond indicates cumulative effect estimate and its confidence interval.   [25,29]. Note: Line indicates confidence interval and the green square indicates effect estimate. Black diamond indicates cumulative effect estimate and its confidence interval.
Subgroup analysis by Crocus sativus preparation cannot be performed due to a limited number of trials.
Subgroup analysis by Crocus sativus preparation cannot be performed due to a limited number of trials.
interval and the green square indicates effect estimate. Black diamond indicates cumulative effect estimate and its confidence interval.

Anxiety Level
There was no difference in anxiety level between the Crocus sativus group and the placebo group in 2 trials (SMD: −0.19; 95%CI: −0.76 to 0.38; I 2 statistic = 0%; p = 0.52; 2 studies; 51 participants, low-quality evidence) [26,28] (Figure 11, Table 2). Figure 11. Forest plot for the outcome anxiety level [26,28]. Note: Line indicates confidence interval and the green square indicates effect estimate. Black diamond indicates cumulative effect estimate and its confidence interval.

Depression Level
There was no difference in depression level between the Crocus sativus group and the placebo group in 3 studies (SMD: 0.41; 95%CI: −0.42 to 1.24; I 2 statistic = 82%; p = 0.34; 3 studies; 149 participants, very low-quality evidence) [25,26,30] (Figure 12).  [25,26,30]. Note: Line indicates confidence interval and the green square indicates effect estimate. Black diamond indicates cumulative effect estimate and its confidence interval.  [26,28]. Note: Line indicates confidence interval and the green square indicates effect estimate. Black diamond indicates cumulative effect estimate and its confidence interval.

Discussion
This review aimed to evaluate the effectiveness of Crocus sativus for treating insomnia in adults. We included 8 studies, 2 to 8 weeks long, with 494 participants aged from 18 to 70 years old, and the dosage of Crocus sativus ranged from 0.6 mg to 300 mg. This review showed reduced insomnia severity and increased sleep quality and duration with Crocus sativus compared with the placebos. However, there was no difference between Crocus sativus and placebo for sleep latency, sleep efficiency, sleep disturbances, use of sleep medications, daytime dysfunction, anxiety level, and depression outcomes. Crocus sativus may improve restorative sleep compared with the placebos, but the evidence is very uncertain due to a limited number of trials. No difference was reported in quality of life between the Crocus sativus and placebo groups. The consumption of Crocus sativus has anti-inflammatory effects. It may also be associated with sleep enhancement, as insomnia is associated with increased inflammatory markers. The hypnotic effects in Crocus sativus increase NREM sleep duration and decrease its latency in animal models [11].
The subanalysis for comorbidity could not adequately explain the high sleep quality heterogeneity. The studies among those with diabetes showed consistent findings. In addition to insomnia, most diabetic patients reported factors such as pain in the shin, muscle cramp, increased and decreased blood sugar level, and repetitive urination during the nighttime that affects sleep quality [44].
We comprehensively searched electronic databases to ensure we identified all relevant studies. We also screened the reference lists of all the identified studies to look for any further relevant studies. Two authors reviewed the studies for inclusion and exclusion, extracted data independently, and examined the quality of the studies. The studies included in this review were conducted in various countries, enabling applicability to various settings. However, the applicability of the findings were limited by the duration and dosage of the Crocus sativus administered in the trials.
To reduce publication bias, we searched numerous databases without regard for the publication date and reviewed the reference lists of all relevant articles for additional references. However, we could not confirm whether we had discovered all the pertinent trials in this area. Inadequate trials for each outcome also prevented us from constructing a funnel plot to assess bias and heterogeneity.
To date, limited systematic reviews and meta-analyses have explicitly examined the effect of Crocus sativus on sleep outcomes in human trials. Most research on the efficacy of Crocus sativus has been conducted in animal models. One review that used animal models as subjects concluded that Crocus sativus reduced anxiety and increased sleep time in mice [45], and that it has anxiolytic and hypnotic effects [11]. Studies on human subjects show that Crocus sativus reduced sleep disorders in postmenopausal women [46]. Menopausal syndromes are always associated with symptoms of hot flushes, insomnia, headache, fatigue, depression, and irritability [47]. These symptoms are always associated with hormonal changes during a menopausal state [48]. Thus, the consumption of aphrodite, composed of several plants, such as ginger, saffron, cinnamon, and Tribulus Terrestris improves menopausal symptoms [46].

Implications for Clinical Practice
Presently, moderate-to very low-quality evidence shows that taking Crocus sativus may benefit people with insomnia. There is not much evidence regarding the adverse effects of Crocus sativus on insomnia; thus, the severity of dependency and withdrawal are not known. The findings of the included studies pertain only to short-term use of Crocus sativus (in low doses), with the longest intervention time being eight weeks. This non-pharmacological intervention may reduce the chance of adults with insomnia taking sedative-hypnotic medication, thus reducing dependency and withdrawal symptoms.

Implications for Future Research
There are various methods to quantify sleep parameters, including subjective and objective measures, yet there is no consensus on the best method to use. A specific method or tool for gathering data for future studies is suggested. Though the gold standard and objective measure for sleep assessment is polysomnography, it requires participants to sleep in the laboratory or to be connected to portable polysomnography devices at home. This is impractical and burdens the participants [49]. None of the trials in this review used polysomnography to assess sleep. Alternative tools for objective sleep measures include actigraphy, observation, bed sensors, eyelid movement-and non-invasive arm sensors, a sleep switch, and a remote device [49]. An actigraph device is a non-invasive tool capable of diagnosing circadian rhythm sleep-wake disorders. It is a low-cost and convenient wearable device [50,51]. It does not directly measure sleep, but movement, which is then used to estimate sleep/wake cycles. Only one trial [27] in this review used actigraphy. More studies on humans are necessary. We have assessed adverse events, but very few of the included studies reported sufficient details.

Conclusions
This systematic review and meta-analysis found limited and very low-to moderatequality evidence for Crocus sativus benefitting people with insomnia.