Supplements for mood and stress: what the evidence supports

This hub groups three outcomes measured in people without a diagnosed condition: perceived stress, mood and calmness. Each is a self-report, so a trial's result depends on the questionnaire it used and on how well the placebo group was matched. Where a grade rests on trials in a diagnosed condition, the paragraph names that population and notes that results may not apply to healthy adults.

Perceived stress. The Perceived Stress Scale (PSS) is the most popular instrument for assessing perceived stress [1]. It measures the degree to which situations in a person's life were rated as stressful over the last month; in the version one holy basil trial used, scores range from 0 to 40, with higher scores meaning more perceived stress [7]. Trials also use the stress subscale of the Depression, Anxiety and Stress Scales (DASS) [16] and standardized laboratory stressors such as the Trier Social Stress Test [21].

Mood. The 40-item abbreviated Profile of Mood States (POMS-A) rates current mood on scales including tension, anger, fatigue, vigor, confusion and total mood disturbance [7]. The POMS was the primary outcome in one saffron trial [11], and other mood trials used the Depression, Anxiety and Stress Scales [12]. Positive and negative affect are scored separately on scales such as PANAS [11].

Calmness. Calmness is a single self-rated mood dimension, scored in some trials on the Bond-Lader mood scales [34]. In one Panax ginseng trial, mood, including calmness, was tested on days 1 and 8 of each treatment period, before and at 1, 2.5 and 4 hours after the dose [22].

Cortisol is not the outcome. Cortisol, a hormone released during stress, is a biomarker. In one meta-analysis, ashwagandha lowered cortisol while Perceived Stress Scale scores did not change significantly [6], which is why grades on this page rest on how people felt, not on hormone levels.

Sleep, physical activity and mindfulness training have each been pooled in meta-analyses of healthy people [2][3][4].

Sleep. A preregistered meta-analysis of 154 experimental studies (5,717 healthy participants aged 7-79) found that total sleep deprivation, partial sleep restriction and sleep fragmentation all reduced positive affect (SMD -0.27 to -1.14) and increased anxiety symptoms (SMD 0.57-0.63) [2]. Findings for negative affect were mixed and depended on the type of sleep loss [2].

Physical activity. A meta-analysis in healthy individuals from childhood to old age found a small positive relation between physical activity and subjective well-being (d = 0.360) [3]. Physical activity was more strongly related to positive affect than to cognitive well-being, and was unrelated to negative affect [3].

Mindfulness training. In a meta-analysis of 29 studies (2,668 healthy adults), mindfulness-based stress reduction had a between-group effect size (Hedges' g) of .53 on stress, anxiety and related measures, with large effects on stress and results maintained at an average of 19 weeks of follow-up [4]. Heterogeneity was high, probably because the studies differed in design, protocol and outcomes [4].

Supplement results below should be read against this background.

The notes below, in alphabetical order, summarize the main studies behind each grade.

5-HTP

In 46 healthy undergraduates, a single dose of two 50 mg capsules did not significantly alter mood on the Profile of Mood States compared with placebo [30]. A 12-week trial in 30 cognitively healthy older adults compared 100 mg/day with taking nothing rather than a placebo; Geriatric Depression Scale scores improved within the 5-HTP group by week 8, and there was no effect on its anxiety inventory [31]. See 5-HTP.

Acetyl-L-carnitine

The mood evidence for acetyl-L-carnitine comes from clinical populations. A meta-analysis of 12 randomized trials (791 participants, mean age 54) in people with depressive symptoms found fewer symptoms than with placebo or no intervention, with high heterogeneity (86%), and called for larger trials [27]. Results may not apply to healthy adults. See acetyl-L-carnitine.

Apigenin

The human trials behind this grade tested chamomile (Matricaria recutita) extract, so the grade applies to chamomile extract, not isolated apigenin [25][26]. A meta-analysis of 3 randomized trials found no difference in anxiety (SMD -0.15) [25]. In 57 outpatients with mild to moderate generalized anxiety disorder, a clinical population, an 8-week trial found a greater fall in anxiety ratings than placebo (P = 0.047); results may not apply to healthy adults [26]. See apigenin.

Ashwagandha

In a meta-analysis of 12 randomized trials (1,002 participants aged 25-48), ashwagandha extract lowered stress scores compared with placebo (SMD -1.75), but heterogeneity was high (83.1%) and the certainty of evidence was rated low [5]. A 2024 meta-analysis of 9 trials in 558 people with stress or anxiety found lower Perceived Stress Scale scores and serum cortisol than placebo [37]. A 2025 meta-analysis of 6 trials that used the Perceived Stress Scale (488 participants in the review) found no significant effect (SMD -0.355), even though cortisol fell [6]. A 2026 dose-response meta-analysis of 22 trials reported a significant association between dose and stress scores and called for better trials [38]. Evidence is mixed, and the 2025 review authors called for longer trials [6]. See ashwagandha.

B vitamins

These grades apply to combinations. A meta-analysis of randomized trials giving at least 3 B vitamins daily for at least 4 weeks, partly funded by Bayer HealthCare, found a small benefit for stress (958 participants, SMD 0.23) in healthy and "at-risk" adults [14]. The benefit on low-mood symptom scales did not reach significance (p = 0.07), and there was no effect on anxiety [14]. See B vitamins.

Bacopa monnieri

In a 12-week trial in 72 healthy adults aged 35-60, funded by Hindustan Unilever, 450 mg/day of an extract changed no cognitive measure and showed only a trend toward lower state anxiety [17]. In a 12-week trial in 101 adults aged 40-70 with self-reported memory and attention problems, funded by US Pharma Lab, 300 mg/day of the Bacumen extract reduced self-reported stress reactivity more than placebo (p = 0.03), a secondary outcome [18]. See Bacopa monnieri.

GABA

A systematic review of 14 placebo-controlled human trials of oral GABA, whose first author was employed by Unilever, found limited evidence for stress [32]. A separate review noted that most evidence for a calming effect was reported by researchers with a potential conflict of interest [33]. See GABA.

Holy basil

In a registered 8-week trial in 100 adults with stress and sleep problems, funded by Natural Remedies, maker of the Holixer extract, 125 mg twice daily improved Perceived Stress Scale scores more than placebo (p = 0.003), and hair cortisol was lower at week 8 [7]. The same trial found no between-group difference in POMS total mood disturbance (p = 0.695) [7]. A 6-week trial in 150 adults with general stress symptoms, in which three co-authors worked for Natural Remedies, which developed the OciBest extract, found a larger fall in total symptom scores than placebo [8]. See holy basil.

L-theanine

A meta-analysis of 31 randomized placebo-controlled trials (1,168 participants) found only a modest reduction in acute stress after a single dose in healthy adults, its primary outcome (SMD 0.31), largely influenced by studies at high risk of bias; one author founded a supplement company [9]. In a 28-day trial in 30 healthy adults with moderate stress, funded by Ethical Naturals, Perceived Stress Scale scores fell by 17.98% with 400 mg/day of L-theanine and by 17.88% with placebo [10]. Evidence is mixed. See L-theanine.

Lemon balm

In a crossover trial in 20 healthy young adults, single doses of a lemon balm (Melissa officinalis) extract made by Pharmaton raised self-rated calmness at the earliest time points at the lowest dose, while the highest dose reduced alertness [34]. In a trial in 106 young adults with moderate stress, funded by the ingredient supplier Givaudan France Naturals, a single 300 mg dose of the Zensera extract did not change overall calmness, its primary outcome, though calmness recovered after demanding tasks, which was not seen with placebo, a secondary finding [35]. In a 3-week trial in 100 healthy adults with moderate distress or poor sleep, with product donated by the ingredient supplier Indena, the authors reported improved low-mood, anxiety and stress scores with 400 mg/day [36]. See lemon balm.

Lion's mane

In a pilot trial in 41 healthy adults aged 18-45, 28 days of 1.8 g/day of lion's mane (Hericium erinaceus), with product supplied by the manufacturer, gave only a trend toward lower subjective stress (p = 0.051) [24]. The authors said the small sample means the findings should be interpreted with caution [24]. See lion's mane.

Magnesium

A 2017 systematic review of 18 studies found that every study recruited people already vulnerable to anxiety, that the quality of the evidence was poor, and that no study used a validated measure of subjective stress [15]. In a trial in 264 healthy adults with high stress scores and low blood magnesium, funded by Sanofi, maker of the products tested, stress scores fell by 44.9% with magnesium plus vitamin B6 and 42.4% with magnesium alone, with no difference between arms, and there was no placebo arm [16]. See magnesium.

Mucuna pruriens

As of November 11, 2026, the only stress study found (PubMed search) is in a clinical population: 60 men undergoing infertility screening who had psychological stress took 5 g/day of Mucuna seed powder for 3 months, and the authors reported that stress improved [29]. There was no placebo group; 60 fertile men served as a comparison group [29]. Results may not apply to healthy adults. See Mucuna pruriens.

Panax ginseng

In a crossover trial in 30 healthy young adults, 400 mg of the G115 extract improved self-rated calmness, but only at 2.5 and 4 hours on the first day, and 200 mg slowed a fall in mood [22]. In a separate 8-week placebo-controlled trial in 83 healthy adults, 200 or 400 mg of G115 had no effect on positive affect, negative affect or total mood disturbance [23]. See Panax ginseng.

Phosphatidylserine

In young adults, 300 mg/day for a month was associated with feeling less stressed, but only in those with neuroticism scores above the median [20]. In a 42-day trial in 75 healthy men, funded by Lipogen and Lonza, a daily dose of 400 mg phosphatidylserine plus 400 mg phosphatidic acid changed hormone responses to a laboratory stressor in highly stressed men but had no significant effect on the psychological stress response [21]. See phosphatidylserine.

Rhodiola rosea

The main stress trial randomized 80 mildly anxious adults to 2 × 200 mg/day of the Vitano extract or an untreated control group for 14 days, and found lower self-reported stress, anxiety and other mood ratings [19]. It had no placebo group, was funded by Dr Willmar Schwabe, and its authors said a causal relationship cannot be determined [19]. See Rhodiola rosea.

Saffron

In 128 adults with self-reported low mood and no diagnosis, a 4-week trial funded by Pharmactive, maker of the affron extract, found lower POMS Total Mood Disturbance with 28 mg/day than placebo, but no effect at 22 mg/day [11]. Symptoms related to stress were also lower at 28 mg/day [11]. In a 12-week trial in 202 adults with subclinical depressive symptoms, sponsored by Pharmactive, 28 mg/day improved the primary DASS-21 depression score more than placebo, while other secondary outcomes showed no between-group differences [12]. A publicly funded 6-week trial in 51 healthy adults with subclinical depressive symptoms, with two co-authors affiliated with Activ'Inside, found no significant effect on its primary composite score [13]. Evidence is mixed. See saffron.

Uridine

As of November 11, 2026, the only human mood data found (PubMed search) is an open-label case series of 7 adolescents with bipolar disorder, a clinical population, in which uridine for 6 weeks was associated with lower symptom scores [28]. With no control group, the result cannot separate uridine from expectation or natural change, and it may not apply to healthy adults. See uridine.

Safety questions are covered in Are nootropics safe?, and each monograph above has its own safety and interactions section. To compare every compound and outcome side by side, use the evidence matrix, and see how we grade evidence for what each grade requires.

Recent syntheses point in different directions. The 2026 L-theanine meta-analysis found the acute stress effect modest and largely influenced by high-risk-of-bias studies [9]. For ashwagandha, a 2025 meta-analysis found no significant effect on perceived stress [6], while a 2026 dose-response meta-analysis of 22 trials reported a significant association between dose and stress scores [38]. The 2025 saffron trials split: one sponsored by Pharmactive was positive on its primary outcome [12], and one publicly funded trial with two co-authors affiliated with Activ'Inside was not [13]. Grade changes from new studies are listed on what's new.

What supplements have the most evidence for stress?

Ashwagandha has been pooled in meta-analyses: analyses of 12 and of 9 randomized trials found lower stress scores than placebo, though certainty was rated low in the first [5][37], and a separate analysis of Perceived Stress Scale data found no significant effect [6]. L-theanine [9] and B-vitamin combinations [14] have also been pooled. For holy basil, a registered trial in 100 adults, funded by the extract maker, found greater improvement on the Perceived Stress Scale than placebo [7]. The ranked list above shows the current grade for each compound.

Does lower cortisol mean less stress?

Not necessarily. In an ashwagandha meta-analysis, cortisol fell while Perceived Stress Scale scores did not change significantly [6]. In a phosphatidylserine trial funded by Lipogen and Lonza, cortisol responses changed in highly stressed men but the psychological stress response did not [21].

Do sleep and exercise matter for mood?

In a meta-analysis of 154 experimental studies in healthy people, sleep deprivation, sleep restriction and sleep fragmentation all reduced positive affect [2]. A meta-analysis in healthy individuals found a small positive relation between physical activity and subjective well-being, more strongly with positive affect than with cognitive well-being [3].

Are these supplements safe?

Each monograph linked on this page has its own safety and interactions section, drawn from trials, regulators and safety references. Trials summarized here mostly lasted 12 weeks or less, for example 8 weeks for holy basil [7] and 12 weeks for saffron [12], so they say little about longer use.