A scoping review of over-the-counter products for depression, anxiety and insomnia in older people.

Depression, anxiety and insomnia are common in later life, which often involves challenges such as bereavement, chronic diseases, lack of family support and financial instability [1]. Estimates suggest that 7% of older people have major depressive disorder [2], 10% have insomnia [3] and 14% have an anxiety disorder [4]; and that the prevalence of these is increasing in the UK [5, 6]. These conditions are in turn associated with adverse outcomes, such as reduced quality of life [6] and increased risk of physical frailty [7, 8], dementia [9, 10] and mortality [11], and increased use of health services [12].

Current treatments are not optimal, as antidepressants such as selective serotonin reuptake inhibitors (SSRIs) have limited acceptability in older people [13], with negligible effects and an increased falls risk [14]. Similarly, Z-drugs for insomnia (such as zopiclone) are associated with an increased falls and fracture risk when used long term [15]. Although psychological therapies are effective for older people [16, 17], uptake rates are only half of those expected in the UK [18], with multiple access barriers, including local availability and perceptions that older people do not want them [19]. Further barriers exist for internet-based psychological approaches, such as lack of awareness of what is available, limited technological skills and preferences for face-to-face contact for mental health problems in order to build trust [20, 21]. For mild-moderate depressive symptoms and insomnia, evidence suggests older people prefer to self-manage [13, 22], with less surety on managing anxiety [23]. This may be prior to or whilst seeking GP support [24].

Using over the counter products (OTCs) with a claim for treating or preventing depression, anxiety, and insomnia, such as St John’s Wort (Hypericum perforatum L.) or antihistamines, is a key part of this self-management. UK sales of OTC products were £3.2 billion pounds in 2022, with £461 million spent on vitamins and minerals and £63 million on sleep-related products [25, 26]. In a US cohort study of older people, 38% used OTC medication and 64% used dietary supplements for any reason, with 47% of those taking multiple supplements simultaneously [27]. Specifically in insomnia, another survey found 21.9% older people used an OTC medication to improve sleep and 12.5% used herbal medicines, in comparison to 8.3% using prescription medication [28], but did not specify the products. Other surveys of older people focus on products taken rather than for which condition(s) products are taken. From these, commonly used products with a potential indication for mental health include valerian, Nytol herbal (hops, valerian, and passionflower), cod liver oil, multivitamins, vitamin D, gingko (Ginkgo biloba L.), turmeric (Curcuma longa L.), black seed oil (Nigella sativa L.) and cinnamon (Cinnamomum spp.) [29–32].

In light of this, appropriate OTC products with a reasonable evidence base are a potential avenue for reducing depression, anxiety and insomnia symptoms in later life, and if used effectively, could reduce the need for healthcare professional support. However, despite their commonality, OTC product knowledge is limited in older people [33] and pharmacists [34]. Although some OTC products have been systematically reviewed previously, e.g. melatonin [35], omega-3s [36], vitamin D [37], probiotics [38] or St John’s wort [39], many focus on general adult populations and do not compare across products. It is particularly important to assess products in older people, who may be taking other medications or have comorbid conditions, as surveys suggest concurrent use of OTC medication, herbal medicines, prescription medication and dietary supplements are common [29, 40].

Therefore, we aimed to understand the types and characteristics of oral OTC products (including medication, herbal medicines, homeopathic products and dietary supplements) evaluated in trials for reducing symptoms of depression, anxiety and/or insomnia in older people, to identify potential gaps and challenges in product use.

We carried out a scoping review following the Joanna Briggs Institute (JBI) Manual for Evidence Synthesis [41]. A review protocol was developed and uploaded to the Open Science Framework (https://osf.io/rkm57/) before study commencement. The review was performed as part of a larger review focussing on all ages for depression, anxiety, and insomnia. The present review summarises the studies for older adults (60+); for those aged 18–60 separate reviews are being carried out.

After pilot searches in Medline and Embase to identify suitable keywords and subject headings, we carried out a comprehensive search of CENTRAL, MEDLINE, EMBASE, PsycInfo, CENTRAL and AMED (inception to 19th December 2022), combining terms for (1) OTC products, (2) depression, anxiety, and insomnia, and (3) clinical trial terms, using a validated filter where possible (see Appendix A). General OTC product terms were used as there are no established widely applicable product lists. We also searched the reference list of 10% of reviews (n = 200), intending to search more if a large number of relevant new studies were identified, but only two further studies were located. We citation tracked all included studies and followed up all identified protocols and conference abstracts.

Studies were selected based on eligibility criteria in Table 1. In line with scoping review methods, we also outline the context and key definitions in Table 2.

We included studies of all languages and screened using the English abstract and basic translation tools for full texts (e.g. Google Translate). For eligible non-English papers (n = 3), one was extracted by a multilingual team member (MH) and the other two were translated using Google Lens, with a multilingual colleague checking key details where uncertain.

Identified records were deduplicated using EndNote (Version X9.3.3) [43]. Given the large volume of hits (n = 15,539), five reviewers independently screened a total of 10% titles and abstracts in pairs (RF and SU, SM and AS, SM and VT) using Rayyan [44], and met to discuss and resolve agreements and iteratively refine inclusion criteria as per scoping review methods. When agreements were finalised, each reviewer single screened approximately 20% of the remaining hits, with maybes reviewed by a second reviewer. The same approach was adopted for full text screening, using MS Excel.

SM extracted data using a predefined data extraction form in MS Excel, developed for this study and piloted on two studies. Extracted data included study details, country and setting, sample size, target condition(s), participant characteristics, product type and characteristics, outcomes measured and results. Following data extraction, we descriptively summarised data by condition and product in tables and figures. As quality assessment is not usually part of scoping review methods [41] and our aim was to scope the literature rather than provide a definitive conclusion on the strength of evidence for each product, we did not perform quality assessment.

Two Patient and Public Involvement (PPI) contributors (CV and TC) were involved in conceptualizing the idea, finalising the protocol, choice of products to be included, outcomes to focus on and dissemination.

Out of 23,933 records identified in the overarching review, we screened 15,339 titles and abstracts after deduplication and 1346 full texts, and included 431 papers (see Fig. 1). Reference list screening and citation tracking yielded two additional articles. Fifty-seven papers (47 journal articles of 46 studies and 10 trial protocols) focused on older people (aged 60 + years) were found and are summarised in this paper.

Studies were from a range of countries, chiefly Iran (n = 8) [45–52], Italy (n = 7) [53–59], the United States of America (USA, n = 5) [60–64], and China (n = 4) [65–68]. Three were multi-country studies (USA and United Kingdom (UK) [69], Germany and Austria [70], and France and Israel [71]). Fifty-seven studies were published in English, two in Chinese [67, 68], and one in German [72]. Thirty-five studies were single centre and 11 multicentre [46, 58, 59, 69, 70, 73–75]. Most were parallel RCTs, with eight crossover trials [60, 63, 64, 70, 74, 76–78], and sample sizes ranged from 12 to 930. Thirty-nine studies were double-blind, with one single-blind study [51], two unclear [46, 67], and five did not report blinding information [52, 57, 68, 79, 80]. There were few noticeable publication trends over time.

Most studies included both men and women, with five focusing exclusively on either male (n = 1) [47] or female (n = 4) [54, 56, 64, 70] participants. Mean age ranged from 60 to 84 years. Only two studies reported ethnicity information [65, 73]. Twenty-eight studies included participants with no particular comorbidity, and tended to exclude those with other psychiatric disorders, substance abuse, cognitive impairment or significant medical conditions (e.g. cancers, stroke, autoimmune disease). Nineteen studies (41%) included older people with comorbid conditions, including Parkinson’s Disease [48, 81–83], cardiovascular disease [52, 71, 84], Alzheimer’s disease and/or dementia [50, 58, 62, 69], prostate and colorectal cancer [47, 51], a sample with a range of comorbidities [85, 86] and one each of obstructive sleep apnoea [68], haemodialysis patients [75], benzodiazepine withdrawal [87] and post-ischaemic stroke [65].

In most studies, inclusion criteria limited the use of additional depression or insomnia medications (particularly hypnotics), apart from benzodiazepines [61, 76, 87], anti-anxiety or insomnia medication taken for at least one month prior to starting the study [59], venlafaxine [65], variable antidepressants [82] and citalopram [86]. Three trials reported continuing other medication for comorbidities, including chronic heart failure treatment [52], atenolol [71], and acetylcholinesterase inhibitors with or without memantine [69]. Three further studies had trial arms including both an OTC product and a non-pharmacological approach, including a continuous positive airway pressure machine [68], a yoga programme [79] and various lifestyle programmes [77]. Of the studies with concomitant medication, two reported on drug interactions, with no significant interactions [69, 71]. Additional medications taken simultaneously for other conditions were rarely reported unless part of the inclusion criteria.

Included studies focused on insomnia (24/46, 52%), depression (20/46, 43%) and multiple conditions (one all three, one depression plus anxiety) (Fig. 2). No studies focused solely on anxiety.

To assess depressive symptoms at baseline, studies most commonly used the Diagnostic and Statistical Manual of Mental Disorders 4th Edition, Hamilton Depression Rating Scale (HAM-D), and Geriatric Depression Scale (GDS). To assess insomnia, scales such as the Pittsburgh Sleep Quality Index (PSQI), Sleep History Questionnaire, or Insomnia Severity Index (ISI) were used, with a variety of cut-offs. To measure depression as outcome, 10 studies used one outcome measure and 10 studies used multiple measures. The most common measures were the HAM-D (n = 10), Beck Depression Inventory (n = 5), Clinical Global Impression scale (n = 5), GDS (n = 4) and Self-rating Depression Scale (n = 3), with other measures used in two or fewer studies (see Table 3). To measure insomnia as an outcome, four used a single measure and 20 used multiple measures. These were mainly PSQI (n = 10), actigraphy (n = 7), sleep diaries (n = 7), polysomnography (n = 6), Leeds Sleep Evaluation Questionnaire (n = 5) and the ISI (n = 3). Other measures were used in two or fewer studies (see Table 4).

Most studies (n = 37/46) evaluated dietary supplements (Fig. 2). Products were given for between two weeks and 12 months and were predominantly used alone (n = 39), with nine studies evaluating products alongside another therapy [52, 65, 68, 69, 77, 79, 82, 86, 87]. Thirty-three products were compared to a placebo, eight to an active treatment [48, 51, 52, 59, 65, 70, 72, 80], three to both placebo and active drug [58, 82, 86], one to no treatment [49], and one to yoga [79].

For depression, 14 trials evaluated five dietary supplements [46, 50, 51, 54, 55, 57–59, 77, 82–84, 86, 88] (Table 3). Omega-3 fatty acid capsules were most frequently evaluated (n = 7 studies, sample sizes 46 to 271), either in the form of long-chain polyunsaturated fatty acids (eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) [53–55, 83, 84]) or fish oils [50, 82]. Where reported, EPA doses ranged from 180 mg–2 g per day and DHA doses ranged from 120 mg–1 g per day, with supplements taken for between two weeks and six months, indicating a wide dosage range tested. Six out of seven studies comparing omega-3 fatty acids to placebo found a significant improvement in depression symptoms assessed by validated questionnaires, including in people with Parkinson’s disease [82, 83] and dementia [50] but not in people with chronic heart disease [84]. Out of the four studies reporting safety data, two found no side effects and two reported mild gastrointestinal side effects and headache [50, 55].

L-acetyl-carnitine was tested as 3 g/day for seven weeks and 1.5 g/day for 60 days [57, 59]. Despite small studies (n = 80 and n = 28), both were effective compared to fluoxetine and placebo, respectively, with fewer adverse events than fluoxetine and similar events to placebo [57, 59]. Vitamin D3 showed mixed effects across two studies compared to placebo [46, 88], which may be attributable to large dosage differences (50,000IU weekly vs 1200IU per day), with no serious side effects reported. Vitamin B12 and folic acid effectively reduced depressive symptoms compared to placebo [77], but a combination of B6, B12 and folic acid showed no additional effects alongside citalopram in people with comorbidities [86]. 5’-Methyltetrahydrofolic acid (50 mg/day for 8 weeks) showed no difference to Trazodone in people with dementia, with no side effects noted [58].

Six trials (sample sizes 60–161) evaluated six herbal products [49, 65, 70, 72, 79, 80], including 4 g ashwagandha (Withania somnifera (Linn) Dunal) and vacha (Acorus calamus L.) [79], Saraswata Churna (a powdered mixture of 13 herbs, 3 g/day) [80], gingko (120 mg/day Gingko biloba L.) as an adjunct to venlafaxine in people with ischaemic stroke [65], lavender tea (Lavandula, species not reported) [49], St John’s Wort (Hypericum perforatum L. 400 mg/day for six weeks) alone [70], and St John’s Wort + valerian (Valeriana officinalis L.) for 6 weeks [72]. All showed positive effects, four compared to prescribed medication, one to yoga [79] and one to no treatment [49]. Lesser and milder adverse events were reported for gingko compared to venlafaxine [65], and slightly fewer adverse reactions were reported for St. John’s Wort than fluoxetine [70]. No side effects were noted for lavender and sarasvata churna [49, 80]. No trials were found for medication or homeopathic medicines.

Melatonin (n = 17 studies in 18 publications, samples sizes 22 to 930) was the most commonly evaluated dietary supplement, over time periods of 1 week to 12 months. Sustained-release forms were assessed in 10 studies (mostly 2 mg, but varying from 0.5 to 4 mg doses taken before bedtime), and nine showed effects on one or more sleep outcomes vs. placebo [63, 69, 71, 81, 85, 89, 90], whilst one showed no effects [62] and one only measured safety outcomes [87]. Immediate release melatonin was evaluated in five studies (0.5-5 mg before bed), three of which found no effects [62, 78, 90] and two found effects [63, 75]. One evaluated a mix of immediate and sustained release melatonin at two dosage levels and found no effects vs. placebo [61]. Four did not report the form and used doses of 3-6 mg, and found positive effects vs. placebo [67, 76] and better or similar effects to prescribed drugs [48, 51]. Effects were found in both general older populations and people with cardiovascular disease [71], Parkinson’s disease [81] and on haemodialysis [75], with mixed effects in Alzheimer’s disease [62, 69], and similar effects to Zolpidem in people with colorectal cancer [51]. Thirteen out of 18 studies reported safety data and indicated melatonin was well tolerated in older people with no or only mild AEs and few SAEs [48, 51, 62, 69, 71, 76, 78, 85, 89].

Three other dietary supplements were evaluated. Magnesium (500 mg/day for 8 weeks) reduced insomnia compared to a placebo in 46 people (no safety data) [45]. Tart cherry juice (8oz twice/day for two weeks) reduced insomnia severity, but had no effect on other sleep parameters and did not report safety [60]. Pear pulp with added magnesium, melatonin and zinc reduced all insomnia-related parameters compared to pear pulp alone and was tolerated well [56].

For herbal products, Valerian (Valeriana officinalis L. root 100 mg as a soft gel) was ineffective compared to placebo, with no difference in side effects [64]. Melissa officinalis L. syrup (12 ml/day) showed similar effects to alprazolam, and had significantly shorter time taken to fall asleep, but did not report safety [52]. SurAsleep (a mixture of herbs and dietary supplements) was taken for 12 weeks and how effects vs. placebo in 100 people, with no significant adverse events reported [66].

For OTC medication, 1000 mg paracetamol at bedtime for three weeks did not improve sleep compared to a placebo in 61 people, and no safety data were reported [74].

Melatonin (3 mg) was also evaluated for depression, anxiety, and insomnia in 63 people [47]. It significantly improved sleep scores but not depression or anxiety, but excessive daytime sleepiness was a side effect [47]. A second study evaluated rhodiola (Rhodiola rosea L.) for depression and anxiety symptoms in 90 people with comorbid obstructive sleep apnoea [68]. When used with a continuous positive airway pressure (CPAP) machine, rhodiola reduced depression and anxiety, but not alone. No safety information was provided.

No economic evaluations of OTC products were found.

We identified 10 protocols without a related publication – eight were listed as completed and two are ongoing. The two ongoing studies are targeted at insomnia and depression in older people and evaluate a dietary supplement containing green tea, chicory and collagen, and Tiryaqe wabai (Unani herbal preparation). Three completed trials were targeted at depression, evaluating vitamin C, omega-3 fatty acid, and a combination of folic acid plus omega-3 fatty acids. Four protocols evaluated the effects of melatonin (n = 3) and saffron (n = 1) for insomnia (Appendix B). One trial evaluated the effects of prebiotic and probiotic supplements for anxiety.

In this scoping review, we found that surprisingly few OTC products were investigated for depression, anxiety or insomnia in older people – only 46 studies of 21 products were found – and these were mostly tested in those without comorbidities, or with a single specific comorbidity. Given 65% of people aged 65–84 have multiple long term conditions, rising to 82% in the over 85s [92], most trials have been conducted in unrepresentative populations. Few products had multiple trials, apart from omega-3 fatty acids (n = 7), which had more substantive evidence for depressive symptoms, and melatonin (n = 17) which had more substantive evidence for insomnia. Most products had a good safety profile where this was reported, but 13 studies did not report safety data. There was a clear gap in the evidence base in products for anxiety in older people, OTC medications such as antihistamines and homeopathic products.

The products with the most trials do represent some of the most popular products. Omega 3s are commonly used by older people [27, 33]. Systematic reviews of omega-3s for depression in older people find that effects are likely to be dose-dependent and consistent across physical health conditions, and were from good quality studies [36, 93]. A recent meta-analysis of melatonin showed improved sleep quality in a meta-analysis of 23 trials, of mostly good quality, with no differences for people aged over 65 [94]. Although not available OTC in the UK, US melatonin sales to all ages almost tripled between 2016 and 2020, suggesting it is popular [95].

However, there were clear discrepancies between other OTC products tested and used by older people. OTC antihistamines are commonly used to induce sleep in the UK [96], but no studies were found in older adults. One crossover trial located separately found diphenhydramine for two weeks reduced number of awakenings in 25 older people, compared to placebo, but no other sleep parameters, and was less effective than temazepam [97]. Positive effects were found for 50 mg for 1–2 weeks in younger people but greater adverse effects than placebo and a valerian-hops combination [98, 99]. Our review found one paracetamol trial – whilst not indicated for sleep, the authors’ rationale for evaluating this was observing patients self-medicating using this for sleep in geriatric clinical practice [74].

Other commonly used herbal supplements requiring further research in older people include Valerian, Nytol herbal (hops, valerian and passionflower) and St John’s Wort [29], which have little evidence in older people. Products for anxiety in later life are a particularly large gap in the literature requiring further research. Later-life anxiety is less well understood than depression, but is common and is associated with adverse outcomes such as cognitive impairment [9]. Promising anxiolytic herbal products identified in a network meta-analysis of herbal products in younger age groups include Silexan (a lavender oil preparation), kava, passionflower, saffron, gingko and ashwagandha [100]. These should be evaluated in older samples. At present, further comparative reviews of OTC products for older people are unlikely to yield much useful information given the paucity of studies, and further primary studies are needed on promising products.

People aged 65 to 85 are typically living with on average 2.6 long-term conditions [92], which are likely to require medication. Polypharmacy is extremely common and increases the risk of drug-related harm [101]. Future trials need to take a more pragmatic approach and include older people with comorbidities and/or taking prescription medication [102]. Only nine studies evaluated products as adjunct therapies. Real-life supplement use is also likely to involve taking multiple OTC products simultaneously [27]. Whilst some trials evaluated products containing multiple ingredients, further evaluation of combinations are needed. Few surveys assess older people’s usage of OTC products for specific problems; preparatory work should determine what is commonly used prior to new trials.

Financial aspects of OTC product usage are also rarely considered, despite costs for individuals. Our review found no studies on this. It is unclear if OTC product usage reduces healthcare usage through better self-management, reduced medication costs, and reduced risk of adverse events; increases healthcare usage through potential risk or interactions; or has no impact. As UK policy is directing more people towards self-management and seeking pharmacy support [103], this remains an avenue for further exploration.

Strengths of this review include following established guidelines [41], a thorough and iterative search strategy and inclusion of all languages. Due to the large volume of hits, we were only able to dual screen 10% titles, abstracts and full texts. It was particularly challenging to define OTC products in a global context, and it is possible some promising products were missed. Herbal products and dietary supplements are also not necessarily consistent across supplier and product, or may straddle the definition of food and supplements (e.g. tart cherry juice). We employed an inclusive approach in this review to encompass a wide range of products, as advised by our public contributors, but different implications may arise from this review in different contexts.

UK NICE guidelines do not currently recommend OTC products for insomnia or depression [104, 105]. For older people, based on the current review this recommendation seems reasonable. However, there are obviously high OTC product usage levels and some which have a more substantial volume of evidence. Our review suggests that future primary studies should focus on widely used products, with a more pragmatic approach to testing, including populations similar to those who would use them in practice. There is a particularly large gap in evaluating products for anxiety in older people; future research should prioritise this.

This study is funded by the National Institute for Health and Care Research (NIHR) School for Primary Care Research (project reference 635). The views expressed are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care.

RF, KW, MH, VT, CV, SKB, TC, JCB and CB conceptualised the initial idea and applied for grant funding. RF, SM, VT, SU and AS developed search strategies and carried out screening of titles, abstracts and full texts. SM extracted data and provided an initial draft of results. TC and CV provided Patient and Public Involvement input at all stages. RF led writing the manuscript and all authors reviewed the manuscript.

This study is funded by the National Institute for Health and Care Research (NIHR) School for Primary Care Research (project reference 635). The views expressed are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care.

The data extraction form upon which the review results are based can be obtained from the authors upon reasonable request.

Not applicable (review).

Not applicable.

The authors declare no competing interests,

The data extraction form upon which the review results are based can be obtained from the authors upon reasonable request.