Natural and herbal remedies, also known as “complementary” or “alternative” medicines (CAMs), have grown tremendously in popularity over the past two decades, becoming a major component of health care and general wellness in the United States and worldwide. Many people certainly benefit from them, often in cases when conventional treatments have failed or caused side effects. A 2007 National Health Interview Survey found that 38% of adults and 12% of children had used CAM practices and products in the past year, representing about $33.9 billion in out-of-pocket costs (
1). Although there is growing evidence of efficacy and safety to support the use of these remedies, it is important for clinicians to be aware of the limitations of the evidence base and to take that into account with all the other factors that contribute to clinical decision making (
2). In psychiatry, we have about 40 FDA-approved antidepressants on the market, yet their efficacy has been inconsistent (
3), and many treatment responders will eventually relapse (
4). Continued research on natural therapies is called for, partly because they are readily available over the counter and widely used, and also because of their generally good tolerability and safety.
St. John’s Wort
The herbal remedy St. John’s wort (
Hypericum perforatum L. [SJW]) has been used for centuries to treat depression (
5). There are about 40 published clinical trials, including many comparisons with tricyclic antidepressants (TCAs) and selective serotonin reuptake inhibitors (SSRIs), and various systematic reviews and meta-analyses.
Early clinical studies in depressed samples, conducted mostly in Europe, supported SJW as more effective than placebo and comparable to TCAs, particularly for milder forms of depression (
6). Some reports have supported efficacy for seasonal affective disorder (
7,
8) and for menopausal symptoms (
9). Many early studies, however, were limited by short duration and a lack of standardized diagnostic practices or instruments. More recent studies have been more rigorous, including comparisons between SJW and SSRIs. In some studies, SJW performed comparably to fluoxetine (
10) and sertraline (
11). Other trials comparing SJW to sertraline, fluoxetine, and placebo have suggested no advantage for either medication in samples with moderately severe major depressive disorder (MDD) (
12–
15). However, upon closer scrutiny, the data from these studies suggest that remission rates may, in fact, be higher for SJW and that SJW may be more effective for individuals with less severe depression.
Various systematic reviews and meta-analyses of SJW have been published over the past decade. A Cochrane review by Linde et al. (
6) examined 29 trials including 5,489 patients; 18 trials compared SJW with placebo, and 17 compared it with standard antidepressants. Results showed heterogeneity, but overall, SJW was found to be superior in efficacy to placebo and equivalent to standard antidepressants, with better tolerability. A meta-analysis by Ng et al. (
16) examined 27 clinical trials including 3,808 patients, comparing SJW with SSRIs for treating mild-to-moderate depression. SJW and SSRIs had comparable response and remission rates, and SJW had a significantly lower discontinuation rate than SSRIs. The pooled standard mean difference from baseline scores on the Hamilton Depression Rating Scale (HAM-D) also supports SJW as efficacious. A systematic review by Apaydin and colleagues (
17) examined 35 studies comprising 6,993 patients with mild-to-moderate depression. SJW was associated with more treatment responders than placebo and comparable response rates against standard antidepressants, with a milder side-effect profile than the latter. However, the authors cautioned about heterogeneity in the findings, limited data on severe depression, and poor reporting of adverse effects, particularly rare ones. SJW may also be effective in treating menopausal symptoms, as supported by a meta-analysis by Liu et al. (
18).
SJW contains more than 150 chemicals, some of which have been proposed as the main psychotropic ingredients (
19). Hypericin and hyperforin are the best understood of these (
5,
20), and SJW preparations are typically standardized to one of these chemicals. SJW’s mechanism of action is probably multifactorial (
5), involving an interaction with the hypothalamus-pituitary-adrenal (HPA) axis that results in decreased cortisol production (
21). Other proposed mechanisms include decreased serotonin receptor density, decreased synaptic neurotransmitter reuptake, and direct serotonergic activity (
21). SJW also has very slight monoamine oxidase inhibitor (MAOI) activity. Taking SJW does not require the patient to follow an MAOI diet, but combinations of SJW and SSRIs have resulted in serotonin syndrome (
21). It is, therefore, not advisable to combine SJW with SSRIs.
Typical doses of SJW range from 300 to 1,800 mg/day, usually divided on a three-times-daily basis, with 900 mg/day considered as a standard therapeutic dose. Different manufacturing methods may produce variability in efficacy (
21). The most common side effects are dry mouth, dizziness, and constipation (
21). Less common side effects include phototoxicity (hypersensitivity to sunlight), cycling to mania in patients with bipolar disorder, and drug–drug interactions via the liver enzyme CYP-450–3A4 (
21). These interactions may result in decreased activity of warfarin, cyclosporin, oral contraceptives, theophylline, fenprocoumon, digoxin, indinavir, camptosar, verapamil, benzodiazepines, zolpidem, irinotecam, and olanzapine (
22–
27), and possibly others. Therefore, extreme caution is required for patients with HIV who are taking protease inhibitors, as well as for patients with cancer who are receiving chemotherapy and transplant patients taking immunosuppressive drugs (
21).
The evidence thus far supports SJW for treatment of depressive disorders, although it may be less effective in chronic and/or severe depression. The use of SJW also appears to be cost effective, compared with the use of standard antidepressants (
28). More large-scale, controlled trials are needed to better characterize SJW’s place in the psychopharmacologist’s armamentarium.
Omega-3 Fatty Acids
Omega-3 fatty acids are a family of polyunsaturated lipids, including eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), derived primarily from fish and fish oil preparations, and alpha linolenic acid (ALA), which is derived from vegetable sources such as flaxseed oil (
29). Their role in the treatment of mood disorders has been extensively studied over the past two decades, with generally positive, although mixed, results (
29–
35). Fewer clinical trials have examined omega-3s in other psychiatric conditions, including psychotic disorders and developmental disorders (
32).
Most psychiatric investigations into omega-3 fatty acids have typically used EPA or a combination of EPA and DHA, mostly as an adjunct therapy for unipolar depression (
36) but also as monotherapy. Evidence for DHA therapy alone is more limited (
37–
39), despite its important role in the development of the human brain (
40). The optimal dose of omega-3 fatty acids remains unclear. Early studies have used doses as high as 9–10 g/day, but more recent studies suggest benefit with as little as 1–2 g/day (
29–
35). The differential efficacy of EPA and DHA remains unclear, but some recent comparative studies have favored EPA as the stronger antidepressant (
41–
43).
The data overall are difficult to interpret because of heterogeneity among studies, particularly regarding omega-3 preparations, doses, and study design. Several meta-analyses examined their efficacy for unipolar depression, with mixed results (
29–
31,
33,
34). Doses frequently range from 1 to 2 g/day. Use of preparations with ≥60% EPA in combinations (
29) has been recommended. Recent work by Mischoulon and colleagues suggests that omega-3s may be more effective in individuals who are overweight and/or have high levels of inflammatory markers (
42,
43).
Other potential applications include treatment of perinatal depression (
35) and bipolar disorder (
44,
45), although preferably for the condition’s depressive phase rather than mania (
46). Omega-3s have also been studied for treatment of schizophrenia and other psychotic disorders, with less encouraging results (
47,
48), although there is some evidence of preventive effects (
49–
51). Some benefits have been reported in borderline personality disorder (
32,
52). Recent reports support beneficial mood effects in children and adolescents (
53), as well as some benefit in attention deficit disorders (
54,
55). Evidence of efficacy in dementia is not as convincing (
56,
57). They are likely safe for pregnant women, as eating fish is considered safe during pregnancy (
58), and studies of supplements have suggested benefits for the fetus (
59), but caution should still be exercised in this population. These reports collectively suggest a broad range of applications for omega-3s.
The mechanism of action of omega-3s with regard to psychiatric symptoms is unclear but probably multifactorial. Proposed mechanisms include inhibiting G-protein signal transduction, membrane stabilization, and anti-inflammatory effects, among others (
60,
61).
Omega-3 preparations typically contain 500–1,000 mg of omega-3, usually constituting a mixture of DHA, EPA, and other essential fatty acids such as ALA. Preparations of purified individual omega-3 fatty acids are also available. As with most nutraceuticals, there are no published head-to-head studies comparing different brands or preparations of omega-3s, which limits the ability to make specific recommendations.
Omega-3 fatty acids are a promising treatment, given their tolerability (gastrointestinal [GI] distress is the main complaint reported in most studies) and potential benefits in various psychiatric conditions. There have been anecdotal cases of cycling to mania in bipolar patients who were depressed (
40); individuals with bipolar disorder should, therefore, take omega-3s with concurrent mood stabilizers. Previous concerns about elevated risk of bleeding in postsurgical patients have been mostly disproven (
62), despite the fact that caution is often recommended in patients taking anticoagulants (
63). More placebo-controlled studies of effectiveness with large samples, comparisons of safety and effectiveness between the different types of omega-3 fatty acids, and further investigations into the various possible indications and mechanisms of action of the omega-3s are needed.
S-Adenosyl Methionine (SAMe)
SAMe is an intermediate in the metabolism of folate and B12 and donates methyl groups used in the synthesis of brain neurotransmitters such as serotonin, acetylcholine, and dopamine (
64), with potential indirect effects on norepinephrine synthesis (
65). The literature thus far generally supports parenteral (intravenous [i.v.] or intramuscular [i.m.]) and oral SAMe preparations as more effective than placebo and comparable with tricyclic antidepressants at doses up to 3,200 mg/day (
65–
70). There are more than 50 clinical trials of SAMe for depressive disorders, including 17 open ones, 19 randomized controlled trials, and 21 comparisons with other antidepressants. Doses of SAMe have ranged from 200 to 3,200 mg/day (
70). Most comparisons have been with TCAs, and, surprisingly, only one clinical trial has compared SAMe with SSRIs. A study by Mischoulon et al. (
71) with 189 MDD subjects found no advantage for SAMe or escitalopram over placebo but found that male subjects seemed to improve more robustly than female subjects in this sample (
72).
There has been only one major meta-analysis of comparisons of SAMe with placebo published in the past 15 years (
68). This analysis covered 28 of 47 studies. SAMe’s effect was found to be superior to that of placebo for efficacy measures; however, significance was attained only for effect size analysis. The pooled estimate of effect reflected a 5- to 6-point drop on the HAM-D, which represents solid improvement but certainly not full resolution. Comparisons with other antidepressants suggested no significant differences in risk ratios and effect sizes, which supports equivalence between SAMe and other agents, but most of these studies were limited by the lack of a placebo arm. Concerns were raised about publication bias, small samples, and heterogeneity of doses and delivery systems; for these reasons, the authors were cautious in their interpretation of the results. The encouraging findings, overall, support further study, including comparisons of cost-effectiveness, tolerability, side-effect profiles, and patient preference. A recent review by Sharma and colleagues (
70) noted that relatively few new studies of SAMe have emerged since 2002, suggesting that the meta-analysis by Hardy et al (
68) remains highly relevant.
SAMe may be combined with conventional antidepressants, in some cases, accelerating onset of action (
73,
74). Alpert et al. (
75) found a benefit of SAMe augmentation in 30 SSRI nonresponders after 6 weeks of open-label SAMe at 800–1,600 mg/day. Papakostas et al. (
76) followed up this work with a study of 73 SSRI/SNRI (serotonin norepinephrine reuptake inhibitor) nonresponders who were randomized for 6 weeks to SAMe at 800 mg BID (twice a day) or placebo BID; results showed a significant advantage for SAMe. Apart from these studies, there is a surprising lack of augmentation trials combining SAMe with newer agents.
SAMe may be a particularly good choice for people with the methylene tetrahydrofolate reductase (MTHFR) deficiency, as it would theoretically bypass the steps in folate metabolism that are dependent on this enzyme (
65). SAMe may have a faster onset of action than conventional antidepressants, with some patients improving within 2 weeks (
67).
Other potential applications of SAMe include those for depression with comorbid medical conditions such as Parkinson’s disease, osteoarthritis and fibromyalgia, sexual dysfunction, neurocognitive disorders, psychotic disorders, and liver disease (
70), but evidence is less robust in these cases. It also appears to be safe in pregnancy, but evidence is too preliminary to make definitive recommendations (
70).
SAMe is well tolerated and very safe (
65). Its primary reported side effect is GI upset, but other adverse effects may include insomnia, anorexia, dry mouth, sweating, dizziness, and nervousness (
65). Individuals with bipolar depression have reported increased anxiety, mania, or hypomania (
66,
70,
77). SAMe should, therefore, be used with a concurrent mood stabilizer in this population.
While reported doses of SAMe range from 400 to 3,200 mg/day (
70,
71), some patients with depression who were observed in clinical practice may require even higher doses, which appear safe to administer if the patient tolerates the higher doses well. So far, no serious drug–drug interactions have been reported. SAMe appears very promising, and further research will clarify its role in the psychiatrist’s armamentarium.
Valerian
Valerian (
Valeriana officinalis) is a popular herbal sedative and mild hypnotic that has been used worldwide for over 1,000 years (
78). Its soporific effect is likely due to its active ingredients, including valepotriates and sesquiterpenes, which may function similarly to benzodiazepines or barbiturates (
78), with GABA-ergic activity (
78).
There are more than 40 published controlled trials of valerian that include healthy participants and symptomatic individuals (
78–
81). Many studies suggest efficacy comparable with that of benzodiazepines, with fewer side effects and no apparent tolerance (
78). There is also evidence that valerian is beneficial in children (
82) and the elderly (
83–
85), as well as in menopausal women (
86). However, a few meta-analyses and systematic reviews are not very supportive of efficacy (
79–
81). One small study suggests its benefits in treating obsessive–compulsive disorder (OCD) (
87). It may be effective for treating insomnia in postmenopausal women (
86,
88).
One possible reason for the inconsistent findings in clinical trials may be the powerful smell of valerian that can result in unmasking (
78), but this can be overcome by adding some of valerian’s inactive ingredients into placebos, a more recent practice. Recommended doses of valerian are 450–600 mg before bedtime (
78), and there is no apparent increased benefit from higher doses. Valerian appears to promote natural sleep after a few weeks of regular use, rather than treating insomnia acutely (
78).
Valerian is well tolerated. Headaches and GI complaints are the ones most often reported (
78). There is no evidence of a hangover effect in the morning after use. It appears safe in overdose and has shown no significant adverse interactions (
89). Some rare toxic reactions have been reported, including blurry vision, dystonias, hepatotoxicity (
90–
92), withdrawal, and delirium (
93). There is no evidence of dependence or daytime drowsiness, and it appears safe for children (
94) and elderly individuals (
95). Retrospective studies suggest safety in pregnancy, but valerian use in pregnant women appears uncommon. There is some concern about rare outcomes such as malformations (
96); for this reason, its use should probably be avoided during pregnancy. In summary, valerian appears to be a promising hypnotic that decreases sleep latency and improves sleep quality, with potential niches in various patient populations.
Melatonin
Melatonin is a hormone derived from serotonin in the pineal gland that helps to regulate circadian rhythms (
97). It is often used by people who travel across time zones to reset their biological clock and prevent or alleviate jet lag. Melatonin appears to be an effective hypnotic that works fairly quickly upon administration. It may be more effective for generally healthy people with insomnia due to circadian disturbances (
98,
99) as opposed to those with diagnosed psychiatric disorders. Melatonin may function by interacting with the suprachiasmatic nucleus to reset the circadian pacemaker and attenuate the alerting process (
100), and it may have a direct sedative effect (
101).
There are about 20 clinical studies of melatonin for insomnia, including some with children and elderly individuals. Studies in children with sleep disorders (
102) and neurodevelopmental disorders (
103) have produced encouraging results.
A recent meta-analysis by Auld and colleagues (
104) examined 12 double- or single-blind randomized and controlled studies. Results supported significant advantages for melatonin over placebo at reducing sleep onset latency in primary insomnia, delayed sleep-phase syndrome, and regulating the sleep–wake patterns in patients who are blind. A meta-analysis by Ferracioli-Oda et al. (
105) of 19 trials found improvement in sleep onset latency, total sleep time, and overall sleep quality. McCleery et al. (
106) found no benefits of melatonin for sleep disorders in patients with dementia. Likewise, minimal benefits were found for treating patients with Parkinson’s disease (
107). Other meta-analyses and reviews also offer mixed results regarding the degree of efficacy of melatonin versus placebo (
108).
Recommended doses range widely, from 0.5 mg/day to 10 mg/day (
109), with commercial preparations reflecting this range. There is a long-acting version of melatonin dosed at 2 mg/day that has been shown to be effective, particularly in older individuals (
110–
113).
Side effects from melatonin appear to be few and benign. High doses may cause daytime sedation or confusion (
114). Serious adverse effects, although rare, may include decreased fertility and sex drive (
115), hypothermia (
116), and retinal damage (
117,
118). Because of potential interactions with the HPA axis and thymus gland, melatonin may cause immunosuppression and should, therefore, be used with caution in individuals taking steroids or those who are HIV positive (
119,
120).
In summary, melatonin is a promising hypnotic, generally accepted as safe and effective (
121). Apart from the aforementioned concerns about immunosuppressants, there appear to be no adverse interactions with other drugs. It is likely safe to combine with most prescription psychotropics. Because of encouraging data in children with sleep disorders (
102) and neurodevelopmental disorders (
103), as well as in older patients (
110–
113) these may represent particularly good niches for melatonin, since stronger sedative-hypnotics could be avoided in these vulnerable populations.
Ginkgo Biloba
Ginkgo biloba, the seed from the ginkgo tree, has been used in traditional Chinese medicine for thousands of years (
122). Ginkgo’s primary indication is for improving cognitive function (including memory, abstract thinking, and psychomotor function) in individuals with organic brain diseases such as Alzheimer’s dementia.
Over 30 placebo-controlled double-blind trials in populations with dementia have been published (
122). These trials suggest that dementia improves with ginkgo treatment. However, the standards for testing the efficacy of nootropic drugs have changed over the years. Early studies examined cognitive symptoms but not the patient’s functioning in daily activities and need for care (
122). Many meta-analyses and systematic reviews have suggested its efficacy (
123–
127).
Zhang and colleagues (
125) recently reviewed ten systematic reviews that included applications of ginkgo in Alzheimer’s disease (AD), vascular dementia (VaD), mixed dementia, and mild cognitive impairment (MCI). Overall ginkgo demonstrated a dose-dependent improvement in cognition, neuropsychiatric symptoms, and daily activities, with optimal results at doses of 240 mg/day. Tolerability was comparable to that of placebo, and its application for AD did particularly well, with fewer incidences of vertigo, tinnitus, angina pectoris, and headache.
A similar review of 12 reviews by Yuan and colleagues (
126) found varied quality of the studies but generally supported the efficacy of ginkgo compared with placebo regarding cognitive performance, activities of daily living, and clinical global impression, also when dosed at about 240 mg/day, with good tolerability and safety. Limitations of studies included insufficient evidence for periods of less than 22 weeks. The authors also pointed out that doses below 200 mg daily are less likely to be effective.
A meta-analysis by Hashiguchi et al. (
127) examined nine of 13 studies lasting 12 to 52 weeks with doses >120 mg/day in 2,381 patients. Ginkgo outperformed placebo in seven of the nine studies. The advantage was maintained when examining by type of dementia, including AD, VaD, and combined AD and VaD. Doses of 240 mg/day were deemed the most effective, with good tolerability.
Gingko has been compared with synthetic nootropic drugs. Cholinesterase inhibitors are somewhat more effective but not as well tolerated, and may be combined with ginkgo (
128–
132). Ginkgo’s strongest advantage may lie in its lower incidence of side effects (
133); consequently, many physicians have favored ginkgo over the synthetic nootropics as a first-line treatment (
134). Ginkgo appears to have no clear preventive effects on dementia (
135).
Gingko contains many active components, such as flavonoids (quercetin, kaempferol, and isorhamnetin) and terpene lactones (ginkgolides, bilobalide, and ginkgolic acids) (
122). It appears to have multiple mechanisms of action. It is thought to stimulate and protect functional nerve cells from hypoxia and ischemia and free radical damage (
136), and it may stabilize neuronal membranes (
122). Other functions may include preservation of muscarinic choline receptors and alpha-2 adrenergic receptors and promotion of choline uptake in the hippocampus (
122).
The suggested dose of ginkgo is 120–240 mg/day, dosed two to three times a day. Assessment of its full effect may require at least one year (
122,
123). Long-term benefits are unclear. Alzheimer’s dementia may respond better than VaD (
123). Side effects may include mild GI upset, headache, irritability, dizziness, or allergic reactions. There are no established interactions with other drugs (
123,
137), but caution is recommended with patients who have bleeding disorders or are taking anticoagulant medications such as coumadin, because ginkgo inhibits platelet-activating factor (PAF) (
136,
137). Nonetheless, a recent meta-analysis of 18 trials (
138) including healthy volunteers and patients with dementia, peripheral artery disease, or diabetes mellitus did not find any increased risk of bleeding, based on hemostatic outcomes.
Some studies have suggested that ginkgo may alleviate antidepressant-induced sexual dysfunction (
139,
140). In an open trial of ginkgo in 63 patients with antidepressant-induced sexual dysfunction (from SSRIs, SNRIs, TCAs, and MAOIs) (
140), 91% of women and 76% of men reported improvement in all aspects of the sexual cycle with doses of 60–180 mg BID. On the other hand, an open trial by Ashton and colleagues (
141) found limited effect on antidepressant-induced sexual dysfunction in a small sample of 22 subjects. Wheatley (
142) likewise found no significant advantage to ginkgo over placebo in a small study. A controlled study by Kang and colleagues (
143) found benefit with ginkgo, but a high placebo response rate prevented a significant difference after 8 weeks. A study of women with sexual arousal disorder found that ginkgo may have an augmentative effect with sex therapy but not when used alone (
144). The mechanism for improvement in sexual functioning may involve ginkgo’s interaction with PAF, prostaglandins, peripheral vasodilatation, and central serotonin and norepinephrine receptor activity.
In summary, ginkgo appears to alleviate dementia symptoms, or at least to slow down the course of the illness, with a benign side-effect profile. The full extent of its role as a nootropic and as a treatment for sexual dysfunction remains to be clarified. Apart from the risk of hemorrhage in people with bleeding disorders or who take anticoagulants, ginkgo appears to be safe to combine with other medications.