Omega‐3 fatty acid supplementation for depression in children and adolescents

Abstract

Background

Mental health disorders including major depressive disorder (MDD) are well recognized as major contributors to the global burden of disease among adolescents. The prevalence of adolescent depression is estimated to have increased by at least 25% during the COVID‐19 pandemic, compounding the already challenging problem of insufficient mental health service and service accessibility that existed prepandemic.

Omega‐3 polyunsaturated fatty acid (PUFA) supplementation is currently recommended as a preventive treatment for depression in high‐risk adults as well as a second‐line monotherapy for adults with mild to moderate MDD, and adjunctive to antidepressants for adults with moderate to severe MDD. The benefits of omega‐3 PUFA intake on depressive illness have been hypothesized to occur as a result of their effect on neurotransmission, maintenance of membrane fluidity, and anti‐inflammatory action. A comprehensive synthesis and quantification of the existing evidence on omega‐3 PUFA's efficacy in treating depression among children and adolescents is essential for clinicians to provide informed guidance to young people and their families, especially considering the absence of current guidelines for this age group.

Objectives

Primary objective

To determine the impact of omega‐3 PUFA supplementation versus a comparator (e.g. placebo, wait list controls, no treatment/supplementation, or standard care) on clinician‐diagnosed depression or self‐reported depression symptoms in children and adolescents.

Secondary objectives

To estimate the size of the effect of omega‐3 PUFAs on depression symptoms.

To estimate the effect of each type of omega‐3 PUFA (EPA or DHA), compared with placebo, on depression.

To determine if the effect is modified by dosage, format (capsule or liquid), sex, or age.

To determine compliance and attrition for omega‐3 PUFAs as an intervention for depression in children and adolescents.

To determine the safety of omega‐3 PUFAs as an intervention for depression in children and adolescents.

Search methods

We searched CENTRAL, MEDLINE, Embase, three other databases, reference lists of included studies, grey literature, and relevant reviews. The latest search date was 18 May 2023.

Selection criteria

We included randomized controlled trials (RCTs) involving males and females aged 19 years or younger with diagnosed depression comparing omega‐3 PUFA supplementation to placebo, wait list control, no treatment/supplementation, or standard care.

Data collection and analysis

We used standard Cochrane methods. Our primary outcomes were self‐reported depression symptoms and clinically diagnosed resolution of depression. Our secondary outcomes were attrition, adverse effects, and compliance with the intervention. We used GRADE to assess the certainty of evidence for key outcomes.

Main results

We included five trials with 228 participants in our meta‐analysis. All trials used a placebo comparator; intervention duration ranged from 10 to 16 weeks (median: 12 weeks).

Omega‐3 PUFA supplementation compared to placebo may reduce self‐reported depression symptoms, but the evidence is very uncertain (standardized mean difference [SMD] −0.34, 95% confidence interval [CI] −0.85 to 0.17; lower SMD means greater improvement in depression due to omega‐3 PUFA; 5 trials, 185 participants; very low‐certainty evidence).

Omega‐3 PUFA supplementation may have little to no effect on remission of depression symptoms compared to placebo, but the evidence is very uncertain (omega‐3 PUFA versus placebo: 50% versus 48%; odds ratio [OR] 1.11, 95% CI 0.45 to 2.75; 4 trials, 127 participants; very low‐certainty evidence).

Omega‐3 PUFA supplementation may result in little to no difference in attrition (dropouts) compared to placebo (omega‐3 PUFA versus placebo: 18% versus 19%; OR 0.94, 95% CI 0.46 to 1.90; 5 trials, 228 participants; low‐certainty evidence).

Omega‐3 PUFA supplementation may result in little to no difference in adverse effects, with one study reporting more muscle cramps in the fish oil group (13/27 participants) compared to the placebo group (6/29 participants); one study reported more frequent defecation in the omega‐3 group (1/29 participants) and one study identified mild skin rash and unusual/manic behavior in the placebo group (2/27 participants).

None of the included studies reported serious adverse effects.

Authors' conclusions

Based on five small studies, omega‐3 PUFA supplementation may reduce self‐reported depression symptoms, but the evidence is very uncertain. Omega‐3 PUFA supplementation may have little to no effect on depression remission compared to placebo, but the evidence is very uncertain. Omega‐3 PUFA supplementation may result in little to no difference in attrition or adverse effects. The studies observed no serious adverse effects.

This review highlights early‐stage research on omega‐3 PUFA and depression in young people. The evidence on the effects of omega‐3 PUFA supplementation in improving self‐reported depression symptoms or achieving depression remission in children and adolescents is very uncertain. While no harms are evident, more data are needed to confirm potential risks.

Addressing current limitations in the evidence base through the design and conduct of methodologically rigorous studies — larger sample sizes, varied dosages, eicosapentaenoic acid/docosahexaenoic acid ratios, treatment durations, and safety profiles — is crucial to advance our understanding of the role of omega‐3 PUFA supplementation for depression in children and adolescents.

Plain language summary

Is omega‐3 polyunsaturated fatty acid supplementation better than a placebo for treating depression among children and adolescents?

Key messages

– Due to a lack of robust evidence, the benefits and unwanted effects of omega‐3 polyunsaturated fatty acids (which are important fats that you must obtain from your diet) supplementation for treating depression among children and adolescents are unclear.

– Larger, well‐designed studies are needed to give better estimates of the potential benefits and harms of omega‐3 polyunsaturated fatty acid supplementation for child and adolescent depression.

Why is it important to find treatments for children and adolescents with depression?

People with depression have a low mood or lose pleasure or interest in usual activities. It is a long‐lasting (chronic) disorder that is both serious and recurrent (it can return after periods of feeling well). It often begins during childhood and adolescence, affecting up to eight in every 100 children. Unfortunately, response to current treatments for depression in children and adolescents is poor, and relieving symptoms (remission) can be challenging. In recent studies, supplements of omega‐3 polyunsaturated fatty acids have shown promise as a potential treatment for depression in adults. Our body cannot make these beneficial fats, but they are naturally present in oily fish, marine sources, and certain nuts and seeds. They can also be given as daily supplements in capsule or liquid forms (for example, as cod liver oil capsules).

What did we want to find out?

We wanted to find out:

– if omega‐3 polyunsaturated fatty acids supplementation works to decrease symptoms of depression in children and adolescents; and

– if omega‐3 polyunsaturated fatty acids supplementation helps children and adolescents achieve remission from depression.

We also wanted to know if omega‐3 polyunsaturated fatty acids supplementation can cause any unwanted effects.

What did we do?

We searched for studies in which children and adolescents were assigned randomly to two or more treatment groups that compared omega‐3 polyunsaturated fatty acids supplementation with a placebo (pretend treatment with the same appearance), on a waiting list, no treatment/supplementation, or usual care. This is the best way to ensure that groups of people are similar and that investigators do not know who is in which group. We compared and summarized their results, and rated our confidence in the evidence, based on factors such as study methods and sizes.

What did we find?

We found five studies that involved 228 children and adolescents ranging from 10 to 16 years of age with depression. The largest study included 60 people, and the smallest study included 28 people. Three studies were conducted in the USA, one in Israel, and one in Slovakia. Four studies lasted for around 12 weeks and one study lasted for 16 weeks.

We found that omega‐3 polyunsaturated fatty acids supplementation may reduce self‐reported depression symptoms, but the results are very uncertain. If omega‐3 PUFA supplements and placebos were given to a group of 100 children and adolescents with depression, then 50 children receiving omega‐3 polyunsaturated fatty acids would achieve remission and 48 receiving the placebo would achieve remission.

What are the limitations of the evidence?

We are not confident in the evidence because there were not enough studies to be certain about the results, and the studies were very small. To date, there is insufficient evidence to determine if omega‐3 polyunsaturated fatty acids supplementation improves depression for children and adolescents.

How up to date is this evidence?

The evidence is up to date to 18 May 2023.

Summary of findings

Summary of findings 1. Summary of findings table ‐ Omega‐3 polyunsaturated fatty acid (PUFA) compared to placebo for depression in children and adolescents.

Omega‐3 polyunsaturated fatty acid (PUFA) compared to placebo for depression in children and adolescents
Patient or population: depression in children and adolescents Setting: any Intervention: omega‐3 polyunsaturated fatty acid (PUFA) Comparison: placebo
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with placebo	Risk with omega‐3 polyunsaturated fatty acid (PUFA)
Depression symptoms, continuous measure assessed with: self‐reported standardized tools follow‐up: range 10 weeks to 16 weeks	‐	SMD 0.34 SD lower (0.85 lower to 0.17 higher)	‐	185 (5 RCTs)	⊕⊝⊝⊝ Very lowa,b,c	Omega‐3 PUFA supplementation may reduce self‐reported depression symptoms, but the evidence is very uncertain. An SMD of 0.2 represents a small difference between groups.
Depression remission, dichotomous measure assessed with: Children's Depression Rating Scale–Revised (CDRS‐R) scores ≤ 28 follow‐up: range 10 weeks to 12 weeks	48 per 100	50 per 100 (29 to 72)	OR 1.11 (0.45 to 2.75)	127 (4 RCTs)	⊕⊝⊝⊝ Very lowd,e	Omega‐3 PUFA supplementation may have little to no effect on depression remission as a dichotomous measure, but the evidence is very uncertain.
Attrition	19 per 100	18 per 100 (10 to 31)	OR 0.94 (0.46 to 1.90)	228 (5 RCTs)	⊕⊕⊝⊝ Lowf	Omega‐3 PUFA supplementation may result in little to no difference in attrition (dropouts).
Adverse effects	All studies reported no serious adverse effects. 1 study reported more muscle cramps in the fish oil group (13/27 participants) than in the placebo group (6/29 participants). 1 study reported more frequent defecation (2–3 times daily) in the omega‐3 group (1/29 participants). 1 study reported mild skin rash and unusual manic behavior in the placebo group (2/27 participants).			228 (5 RCTs)	⊕⊕⊝⊝ Lowg,h	Omega‐3 PUFA supplementation may result in little to no difference in adverse effects.
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; OR: odds ratio; SMD: standardised mean difference
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
See interactive version of this table: https://gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_427384744860227560.

^a Downgraded two levels as the effect size for depression symptoms (continuous) was affected by the sensitivity analysis removing studies at high risk of bias.
^b Downgraded one level due to substantial statistical heterogeneity (I2 = 63%); inconsistency was judged to be serious.
^c Downgraded two levels for imprecision because of small number of participants and wide 95% CIs.
^d Downgraded two levels as the effect size for depression remission (dichotomous) was affected by the sensitivity analysis removing studies at high risk of bias.
^e Downgraded two levels for imprecision due to the wide 95% CIs, which indicates that the range of possible effects on remission was broad.
^f Downgraded two levels for imprecision because of wide 95% CIs. 
^g Downgraded one level for risk of bias because none of the studies had prespecified analysis plans for adverse effects.
^h Downgraded one level for imprecision because of small number of participants.

Background

Mental health problems, including depression, often start in childhood or adolescence. The pooled prevalence estimate from a meta‐analysis conducted on literature published during the first year of the COVID‐19 pandemic was 20.5%, which is nearly double prepandemic child and adolescent depression rates (Racine 2021). Depression usually emerges after puberty; however, the prevalence of depression symptoms increases around 10 years of age, particularly among girls (Maughan 2013). Treatment options for childhood and adolescent major depressive disorder (MDD) and depression symptoms include pharmacologic, psychological, and combined approaches. Poor response to current treatment and high remission rates indicate a need for new treatments (Brent 2008; Goodyer 2007; Locher 2017; Weisz 2017).

Omega‐3 polyunsaturated fatty acids (PUFA), which are naturally found in fatty fish, marine sources, as well as some nuts and seeds, are an emerging potential treatment for MDD; however, findings from research in adults are mixed. Omega‐3 PUFA supplementation is recommended by the International Society for Nutritional Psychiatry Research (ISNPR) as a preventive treatment for depression in high‐risk adults (Guu 2019). Similarly, the Canadian Network for Mood and Anxiety Treatments (CANMAT) recommends omega‐3 PUFA as a second‐line monotherapy for adults with mild to moderate MDD, or adjunctive to antidepressants for adults with moderate to severe MDD (Ravindran 2016). Additionally, evidence from systematic reviews of omega‐3 PUFA has demonstrated efficacy in treating MDD in adults (Hallahan 2016; Liao 2019; Luo 2020; Mocking 2016; Zhang 2019). However, one Cochrane review found only a small to modest non‐clinical benefit for omega‐3 PUFA as treatment for adults with MDD, highlighting a lack of high‐quality evidence (Appleton 2021). Any beneficial effects of omega‐3 PUFA are likely dependent on dosage and supplement type (Hallahan 2016; Liao 2019; Luo 2020; Mocking 2016).

Without child‐ and adolescent‐specific research on efficacy, supplement use may result in unnecessary exposure to ineffective, improper dosing, unique pediatric adverse effects, or effects on growth and development (Rodriguez 2008). Currently, no guidelines for omega‐3 PUFA exist for children or adolescents with depression. The current review examined the benefits and harms of omega‐3 PUFA treatment for depression in children and adolescents.

Description of the condition

MDD is a common type of unipolar depressive disorder that occurs in childhood and adolescence, characterized by a persistent sad mood, irritability, or loss of interest in activities for at least two weeks (APA 2019). Persistent depressive disorder, which includes dysthymia, is similar to MDD and characterized by less profound though longstanding sadness, with continued symptoms for at least one year (APA 2019). Diagnosis is achieved following clinical assessment. Individuals who exhibit symptoms of depression that do not reach the threshold for MDD, termed 'subthreshold' depression, are also subject to an increased risk of MDD and decreased quality of life (Bertha 2013). Depression is a broad term used to describe these disorders.

Despite its high prevalence and socioeconomic impact, the pathophysiology of depression is not well understood (Deacon 2017). Depression often has biologic, psychological, and social underpinnings (Maughan 2013). Risk factors for depression include familial, genetic, psychosocial, and neural/neuroendocrine risks (APA 2019). During childhood and adolescence, depression has far‐reaching negative implications on physical, emotional, and social development (WHO 2020). The impacts of MDD and its associated consequences may decrease the overall quality of life and health outcomes of affected youth, including increased illness severity, increased risk of comorbidity, hospitalizations, obesity, and premature cardiovascular death (Korczak 2009; Luppino 2010; Osby 2001). MDD affects approximately 1% to 2% of children, and prevalence increases to 5% to 8% with adolescent age (Maughan 2013). Among adolescents, depression is a major cause of disability and can increase the risk of death by suicide, which is the second leading cause of death for people aged 15 to 19 years (WHO 2020).

Description of the intervention

Omega‐3 PUFA are an emerging treatment for depression and other neuropsychiatric disorders. There are three types of naturally occurring lipids, classified by the number of carbon–carbon double bonds present in their fatty acid side chains: saturated, monounsaturated, and PUFA. PUFA are further classified based on the position of the first carbon–carbon double bond site: n‐3 or omega‐3 and n‐6 or omega‐6 (Deacon 2017).

Omega‐3 PUFA are considered essential as they cannot be synthesized in the body, and are naturally found in plant and animal food sources. The main sources of omega‐3 fatty acids are alpha‐linolenic acid (ALA), eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA). ALA is found in plant sources such as canola, hemp, walnuts, and flaxseed (Logan 2004). Within the body, ALA is converted to EPA and DHA (Parker 2006). The conversion of ALA to EPA and DHA is inefficient in humans, with studies suggesting that only 1% to 15% of ALA is metabolized (a process which takes place primarily in the liver) (Harris 2010; Sontrop 2006). Additionally, the extent to which ALA is converted to EPA and DHA varies among individuals, making consistent dosing problematic (Burdge 2006). Dietary sources of EPA and DHA include oily fish such as tuna, salmon, mackerel, or sardines. Dietary supplements may contain combinations of EPA and DHA; however, the most effective dietary supplement preparations for depression in adults have at least 60% EPA relative to DHA (Liao 2019). Fish oils, which contain both EPA and DHA, have been associated with reduced depression risk in observational research. Consuming EPA and DHA directly from foods or dietary supplements is considered the most effective way to increase levels of these fatty acids in the body, with six months being the minimum duration of supplementation for omega‐3 PUFA to equilibrate in the brain (Browning 2012). Other omega‐3 PUFAs, such as docosapentaenoic acid (DPA), heneicosapentaenoic acid (HPA), and the newly discovered eicosatrienoic acid (ETE), are far less abundant in the diet. These omega‐3 PUFAs are not considered essential fatty acids and there exists minimal evidence regarding their relationship with depression symptoms in adults (Kotlega 2020). Some commercial EPA and DHA supplements contain minimal amounts of HPA, ETE, DPA, or combinations of these, and these are not considered to be primary active ingredients. Furthermore, these are not well researched and thus dosage and indications in the pediatric population are outside the scope of this review. EPA and DHA are the only omega‐3 PUFAs included in this review.

How the intervention might work

Omega‐3 PUFA likely act on multiple processes within the body and are necessary for brain function and development. The role of omega‐3 PUFA in the treatment of pediatric depression has been suggested to have the same pathway of action as in adults. The benefits of omega‐3 PUFA intake for depressive illness have been hypothesized to occur as a result of their effect on neurotransmission, maintenance of membrane fluidity, and anti‐inflammatory action (Burhani 2017; Khanna 2019). Omega‐3 PUFA are also crucially important for brain development. Omega‐3 PUFA metabolism is needed for myelination and synaptic pruning, which are core processes during normal pubertal brain development (McNamara 2019). Omega‐3 PUFA may also counteract changes in the hippocampus seen in individuals with depression (Godos 2019; Kang 2013).

Omega‐3 PUFA are essential components of intracellular and neuronal cell membranes, affecting cell membrane integrity and fluidity, and resulting in improved neurotransmission. Membrane organization and the associated assembly of signaling proteins have been associated with the induction of depression‐ and anxiety‐related behaviors (Muller 2015). DHA, in particular, plays a vital role in maintaining cell membrane integrity and fluidity (Parker 2006). An increase in membrane fluidity facilitates transmission by increasing the flexibility of the membrane. Through their effect on the efficiency of membrane functioning, omega‐3 PUFA plays a role in a variety of biologic mechanisms, such as enzyme and receptor activity, ion channel functioning, and the production and activity of neurotransmitters. Changes in membrane fluidity also affect the structure and functioning of proteins embedded in the membrane. This influences the activity of membrane‐bound enzymes (Bowen 2002), the number and affinity of receptors, and the function of ion channels, all of which alter proper neurotransmission. Additionally, the production and activity of neurotransmitters, gene expression, as well as neuroplasticity are also affected through the impact of omega‐3 PUFA on neurotrophin (protein that supports neuronal growth, development, survival, and function) such as brain‐derived neurotrophic factor (Deacon 2017). Omega‐3 PUFA are also directly involved in various neurotransmitter pathways. Omega‐3 PUFA‐deficient diets have been associated with reduced receptor density and disruptions to neurotransmitter activity in serotonergic, dopaminergic, and adrenergic systems compared to people with no deficiency. Disruptions to normal cell signaling, inflammatory processes, and neurotransmitter system activities have been implicated in MDD (Parker 2006).

High levels of inflammation have been associated with the progression and severity of depression among adults and young people (Colasanto 2020; Deacon 2017; Kuhlman 2020). Omega‐3 PUFAs act as anti‐inflammatory agents (Kalkman 2021). Reduced inflammation is associated with better clinical outcomes among people with MDD (Osimo 2019). Chronic stress also activates the immune system, invoking a neuroinflammatory response that releases inflammatory mediators (Joffre 2019). Inflammatory mediators, such as cytokines, influence the activity of the hypothalamic‐pituitary‐adrenal (HPA) axis — an important aspect of the stress‐response system. Some examples of inflammatory mediators are pro‐inflammatory cytokines, including interleukin‐1β, interleukin‐2, interleukin‐6, interferon‐γ, and tumor necrosis factor‐α, which are indicative of the upregulation of inflammatory reactions (Anisman 2002; Deacon 2017). Documented effects of high levels of cytokines also include decreased neurotransmitter precursor availability, altered neurotransmitter metabolism and neurotransmitter transporters (Logan 2004), which may further contribute to depression. The consumption of omega‐3 PUFA has been linked to the reduced production of cytokines (tumor necrosis factor‐α, interleukin 1β, interleukin‐6, C‐reactive protein, serum amyloid A); increased serotonergic and dopaminergic activity; and decreased concentrations of norepinephrine (Deacon 2017). Increased release of anti‐inflammatory mediators, such as eicosanoids, prostaglandins, and leukotrienes, have been attributable to higher levels of omega‐3 PUFA in cell membranes (Gutiérrez 2019).

Why it is important to do this review

This review will fill important gaps in the literature for clinicians, patients, and researchers. There is a growing interest in adjunctive treatments for child and adolescent depression, but recommendations in high‐quality clinical practice guidelines are vague and lack evidence from randomized controlled trials (RCTs), leaving parents and practitioners without clear direction (Campisi 2021a). More importantly, some clinical practice guidelines for adolescent depression currently recommend the use of omega‐3 PUFA supplementation (Kaiser Permanente 2018; Seo 2018). Clinicians cannot advise young people and their families without understanding the evidence base. Thus, a high‐quality synthesis and quantification of the current evidence for the use of omega‐3 PUFA in children and adolescents with depression is required.

Additionally, there are three important patient‐specific factors to consider for the use of omega‐3 PUFA in childhood and adolescent depression. First, child/adolescent onset of MDD has been shown to be a more severe form of illness compared with adult‐onset disease (Korczak 2009), and together with greater atypical symptoms (anger and behavioral problems, as well as irritability as a mood equivalent) suggests there are distinctions between adult‐onset and child‐onset disease that may have treatment implications. This has previously been shown to be the case with pharmacologic treatments (e.g. tricyclic antidepressants) in which RCTs demonstrating benefit for adult depression were subsequently shown to be negative for the treatment of child/adolescent MDD (Hazell 1995). Second, depression is a disease of pessimism and hopelessness, and encouraging children and adolescents with depression to pursue treatments that may not be effective increases their already present thoughts that they cannot be treated or will not get better, and decreases therapeutic alliance for continued care. Third, omega‐3 PUFA supplements are costly, and this may have negative effects on families struggling financially. Parents may choose to pursue omega‐3 PUFA supplements instead of other activities or wholesome foods, for example. Therefore, sound evidence upon which to make a recommendation for omega‐3 PUFA supplementation is critical so that families are able to make evidence‐informed decisions.

Providing high‐quality evidence for effective alternative therapies for depression is of importance as omega‐3 PUFA emerge as an alternative or adjunctive therapy. Although several systematic reviews conducted in adults suggest a positive antidepressant effect of omega‐3 PUFA supplementation (Hallahan 2016; Liao 2019; Luo 2020; Mocking 2016), particularly when used as adjunctive therapy (Liao 2019), one updated Cochrane review found insufficient high‐quality evidence to determine the effects of omega‐3 PUFA as a treatment for adult MDD (Appleton 2021). Furthermore, limited evidence regarding the efficacy of omega‐3 PUFA supplementation for the treatment of depression in children and adolescents currently exists (Lopresti 2015); Zhang 2019 limited studies to those involving omega‐3 PUFA without adjunctive treatment and excluded people with comorbid physical or psychiatric disease (or both). This current Cochrane review implemented rigorous search methods and considered recent trial evidence with wider eligibility criteria to better reflect patient characteristics in the real world.

Objectives

Primary objective

To determine the impact of omega‐3 PUFA supplementation versus a comparator (e.g. placebo, wait list controls, no treatment/supplementation, or standard care) on clinician‐diagnosed depression or self‐reported depression symptoms in children and adolescents.

Secondary objectives

• To estimate the size of the effect of omega‐3 PUFAs on depression symptoms.

• To estimate the effect of each type of omega‐3 PUFA (EPA or DHA), compared with placebo, on depression.

• To determine if the effect is modified by dosage, format (capsule or liquid), sex, or age.

• To determine compliance and attrition for omega‐3 PUFAs as an intervention for depression in children and adolescents.

• To determine the safety of omega‐3 PUFAs as an intervention for depression in children and adolescents.

Methods

Criteria for considering studies for this review

Types of studies

We included all RCTs involving males and females aged 19 years or younger with diagnosed depression comparing omega‐3 PUFA supplementation to placebo, wait list control, no treatment/supplementation, or standard care regardless of their quality. We also included cross‐over and cluster‐RCTs. For cross‐over trials, only data from the first phase will be extracted. We included all published and unpublished protocols and trials that met the inclusion criteria.

Types of participants

Studies with participants of either sex aged 19 years or younger and from any geographical location were eligible for inclusion.

Diagnosis

We included studies that recruited people with depression (regardless of their other symptoms).

• Diagnosed depression (MDD or dysthymia) using validated diagnostic criteria (e.g. Diagnostic and Statistical Manual of Mental Disorders Fourth Edition, Text Revision [DSM‐IV‐TR] (APA 2013), International Classification of Diseases 10th Edition [ICD‐10] (WHO 2016), clinician‐rated Children's Depression Rating Scale–Revised [CDRS‐R] (Poznanski 1996)), or other standardized criteria such as the Research Diagnostic Criteria (Spitzer 1978); or

• Self‐report of clinically significant depression symptoms by standardized questionnaire validated in children and adolescents with depression symptoms. These participants may not have a diagnosis of MDD or dysthymia, but must have displayed symptoms of sadness, loss of energy, change in appetite or sleeping patterns, anxiety, restlessness, feelings of worthlessness, guilt, hopelessness, indecisiveness, or thoughts of self‐harm or suicide prior to randomization (APA 2019). Participants may have had comorbidities (physical or physiologic), but depression must have been the primary diagnosis.

We did not include studies if diagnoses were unclear or investigators used an unstandardized questionnaire or assessment method. We did include RCTs that use a combination of both clinician‐diagnosed depression and self‐reported depression symptoms. In studies where only a subgroup of the participants met the diagnosis criteria, we included the data from that subgroup if the study defined and distinguished the subgroup prior to randomization.

Comorbidities

There is a high incidence of comorbidities in people with depression. We included children and adolescents with other psychiatric diseases, provided they had a primary diagnosis of MDD/dysthymia or their self‐reported depression symptoms suggested clinically significant depression symptoms.

Adjunctive therapy

We included RCTs in which participants were also using adjunctive therapies for depression. There is a high frequency of using adjunctive therapies for depressive disorders. Adjunctive therapy must have been present in both the intervention and control groups.

Settings

We included studies from any setting (inpatient/outpatient or community settings).

Types of interventions

We considered RCTs with placebo control or comparative treatment control that assessed the effects of any omega‐3 PUFA supplementation as sole or adjunctive therapy, regardless of the type of omega‐3 PUFA (EPA, DHA, or a combination of both) as long as the trial administered a standardized dose (in capsule or liquid form). We included any dosage or treatment duration of omega‐3 PUFA supplementation as long as these were clearly specified in the study. Studies that did not provide details regarding the type or dose of omega‐3 PUFA were included in the main analysis but excluded from subgroup analyses investigating the effects of supplementation type or dose.

Types of comparators

All studies must have had a comparator. Possible comparators included placebo, wait list controls, no treatment/supplementation, or standard care. We planned to assess differences in effect size due to the use of different comparators in subgroup analyses.

Types of outcome measures

Primary outcomes

• Depression reported as below.

  • Self‐reported depression symptoms as a continuous measure (score) assessed by any continuous validated measure.

    • We evaluated a change in depression symptoms as the difference when comparing the scores before the intervention to those reported after the intervention.

    • The most commonly used validated rating scales for this age group are the CDRS‐R, the Patient Health Questionnaire for Adolescents (PHQ‐A) (Spitzer 1995), and the primary care version of the Beck Depression Inventory (BDI) (Siu 2016). We also considered studies using other validated scales, such as the Children's Depression Inventory (CDI) (Kovacs 2015). Higher scores on the CDRS‐R, PHQ‐A, BDI, and CDI are interpreted as more severe depression symptoms.

  • Clinically diagnosed resolution of depression, that is, remission, as a dichotomous measure, defined by the absence of depression assessed using ICD/DSM diagnostic criteria, CDRS‐R, or other standardized criteria, through a diagnostic interview or expert opinion (Appendix 1).

Where available, data for all outcomes were recorded at baseline, at the end of the treatment/supplementation period, and post‐treatment/supplementation follow‐up.

Reporting of depression (as primary or secondary outcome by study investigators) was used as an eligibility criterion for including studies in this review.

Secondary outcomes

• Attrition: the number of participants who dropped out of the trials for any reason.

• Adverse effects: the number of participants who experienced any adverse effects or any serious adverse effects. Examples of possible adverse effects included unpleasant taste; bad breath; bad‐smelling sweat; headache; or gastrointestinal symptoms such as heartburn, nausea, and diarrhea (NIH 2018).

• Compliance with omega‐3 fatty acid supplementation: by pill counts, blood tests, or both.

As per our protocol (Campisi 2021b), reporting one or more of the secondary outcomes listed above was not an inclusion criterion for the review. Where a trial (published or unpublished) did not appear to report one of these secondary outcomes, we accessed the trial protocol and contacted the trial authors to ascertain whether they measured the secondary outcomes but did not report them. If relevant trials measured these secondary outcomes but did not report the data at all, or not in a usable format, we included them in the review as part of a narrative summary (see Data synthesis).

Search methods for identification of studies

We identified studies to include by searching databases, clinical trial registries, and grey literature sources.

Electronic searches

We carried out systematic searches across all languages and years on 18 May 2023 in the following electronic databases.

• Cochrane Central Register of Controlled Trials (CENTRAL; Issue 5, 2023); in the Cochrane Library

• Ovid MEDLINE (1946 to 18 May 2023, including Epub ahead of print, in‐process, and other unindexed citations)

• Ovid Embase (1947 to 18 May 2023)

• EBSCO Child Development and Adolescent Studies (1927 to 18 May 2023)

• Ovid PsycINFO (1806 to 18 May 2023)

• EBSCO CINAHL (Cumulative Index to Nursing and Allied Health Literature) Plus with Full Text (1981 to 18 May 2023)

An academic health sciences librarian (GBR) developed the search strategies, and a second librarian reviewed these, using peer review of electronic search strategies (PRESS) guidelines (McGowan 2016). A Cochrane Common Mental Disorders Information Specialist also reviewed the Ovid MEDLINE search strategy. To identify randomized trials in Ovid MEDLINE, we used a modified version of the Cochrane Highly Sensitive Search Strategy for identifying randomized trials in MEDLINE: sensitivity‐maximizing version (2008 revision); Ovid format (Lefebvre 2022). The filter was expanded to include additional terms based on the Cochrane Common Mental Disorders Group core search strategy. We adapted the search strategies for each database using that platform's command language, including text words, controlled vocabulary, and subject headings. The search strategies are available in Appendix 2.

Grey literature

We also conducted a grey literature search. We identified relevant authorities, organizations, and stakeholders using the DuckDuckGo search engine as it does not display personalized search results, thus reducing the risk of bias. We also used the Canadian Agency for Drugs and Technologies in Health (CADTH) Grey Matters Checklist and the Cochrane Handbook for Systematic Reviews of Interventions guidelines to identify supplemental search locations (CADTH 2015; Lefebvre 2022). Details regarding the grey literature searches are available in Appendix 3. On 28 and 29 May 2023, we conducted searches to locate any additional or ongoing clinical trials in the following: ISRCTN Registry; CenterWatch Clinical Trials Listing Service; ClinicalTrials.gov; and World Health Organization International Clinical Trials Registry Platform (ICTRP).

Searching other resources

We searched the reference sections of included studies and previously published systematic reviews, and handsearched relevant journals to locate additional trials. We contacted individual authors who were known experts in the field of pediatric depression and nutritional psychology.

Data collection and analysis

We carried out the review according to the recommendations in the Cochrane Handbook for Systematic Reviews of Interventions (Li 2022). We imported search results into Covidence to facilitate the screening and selection of studies. We used Review Manager for all analyses (RevMan 2024). We recorded the number of search results at each stage of the search and selection process, and produced a PRISMA flow diagram (Figure).

1.

PRISMA flow diagram

Selection of studies

Two review authors (SCC, CZ) independently screened all titles and abstract to identify potentially eligible studies, resolving disagreements with a third review author (DJK). We retrieved all full‐text study reports/publications, and two review authors (SCC, CZ) independently screened these to identify studies for inclusion. At this stage, we resolved disagreements through discussion. We recorded reasons for exclusion at this stage. We identified and excluded duplicates and collated multiple reports of the same study so that each trial, rather than each report/record, was the unit of analysis in this review. We recorded the selection process in sufficient detail to complete a PRISMA flow diagram and Characteristics of excluded studies table (Liberati 2009).

Data extraction and management

We extracted data using a data collection form designed a priori by all the review authors. We used Microsoft Office Excel to design the data collection sheet. Two review authors (CZ, SCC) extracted data from three studies to pilot‐test the form, and made modifications as necessary to reconcile differences and ensure consistency.

Three review authors (SCC, AS, CZ) independently extracted outcome data in duplicate from the included studies. We resolved disagreements by consensus or by involving a third review author (DJK). One review author (SCC) transferred data to the Review Manager (RevMan 2024). We double‐checked that data were entered correctly by comparing the data presented in the systematic review with the data collection form. A second review author (CZ or AS) verified the data entry for accuracy.

We extracted the following study characteristics.

• Background: time period when the study took place, study country or countries, funding source(s), whether it was a published or unpublished study, funding, and conflicts of interest

• Setting: laboratory/community center/clinic

• Methods: study design, depression diagnosis method, description of study arms, unit of allocation, start and end date, the total duration of supplementation, post‐supplementation follow‐up duration, and details of any 'lead‐in' periods.

• Participants: total number, age, sex, baseline severity of the condition, diagnostic criteria, inclusion criteria, exclusion criteria, and withdrawals.

• Intervention group details: number randomized/allocated per group, description of the intervention (including dose and type), duration of supplementation and postsupplementation follow‐up duration, delivery of the intervention, providers and their training, concomitant therapies, and comorbidities

• Comparison group details: number randomized/allocated to a group, description of the intervention, duration of supplementation and postsupplementation follow‐up duration, delivery of the intervention, providers and their training, concomitant therapies, and comorbidities

• Outcomes: measurement tool; total number of intervention, comparison, and follow‐up groups. Where available, we extracted data on all primary and secondary outcomes (clinical assessment of depression (dichotomous)/depression symptoms (continuous), self‐reported depression symptoms (continuous), adverse effects, attrition, and compliance). When reporting depression symptoms, we extracted the mean depression scores at baseline, at the treatment/supplementation end, and at post‐treatment/supplementation follow‐up. Likewise, when depression was clinician‐assessed, we extracted the clinician assessment at baseline, at treatment/supplementation end, and post‐treatment/supplementation follow‐up.

Assessment of risk of bias in included studies

Two review authors (AS, SCC) independently assessed risk of bias using the Cochrane RoB 2 tool (Sterne 2019), as outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2022b) and in Appendix 4. We resolved any disagreements by discussion or by involving another review author (DJK).

We assessed the following primary and secondary outcomes at the end of each trial.

• Self‐reported depression symptoms

• Depression remission

• Attrition

• Adverse effects

Our assessment was based on the following risk‐of‐bias domains.

• Bias arising from the randomization process

• Bias due to deviations from intended interventions (we were interested in quantifying the effect of assignment to the interventions at baseline)

• Bias due to missing outcome data

• Bias in the measurement of the outcome

• Bias in the selection of the reported result

• Overall bias

We used the signaling questions in the RoB 2 tool and rated each domain as 'low risk of bias', 'some concerns', or 'high risk of bias'. We strictly adhered to the RoB 2 tool algorithm when assessing the risk of bias for all studies and summarized the risk of bias judgments across different studies for each of the domains listed for each primary outcome. The overall risk of bias judgment for each outcome was considered as part of the GRADE assessment presented in Table.

Measures of treatment effect

Continuous data

We obtained depression symptoms from each trial using all scales reported. Data, as required for meta‐analyses, included number of participants, mean, and standard deviation (SD) of the score at the end of each trial for the intervention and comparator groups. We used only one scale per trial in the meta‐analysis. As most trials reported the CDRS‐R, we used it when possible. Since all outcomes were reported as a continuous outcome but measured on different scales, we used the standardized mean difference (SMD) to determine the pooled estimate of omega‐3 PUFA effect on self‐reported depression, with 95% confidence intervals (CI).

Dichotomous data

Studies reported data on trial attrition as the number of participants who failed to complete the trial. We collected dichotomous data as the number who dropped out of the intervention and comparator groups and analyzed this data using the Mantel‐Haensel odds ratio (OR) with 95% CIs. Similarly, we collected data on adverse effects as the number of individuals experiencing an event rather than the number of events, but we were unable to calculate the OR with 95% CIs. Further, we recorded depression remission and treatment compliance, where possible.

Unit of analysis issues

Multiple treatment arms

In the case of RCTs with multiple arms, in order to avoid a unit‐of‐analysis error, we kept the arms separate and split the control group. All intervention groups of a multiple intervention study are stated in the Characteristics of included studies table.

Timing of outcome assessment

Where the study reported outcomes at multiple time points, we extracted the time point of the treatment/supplementation at the end of each trial. The impact of supplementation duration and follow‐up duration were not explored through a meta‐regression since the number of included studies was fewer than 10.

Dealing with missing data

We contacted trial authors to request missing outcome data (e.g. when a study was identified as abstract only) since such an omission can potentially contribute to selective outcome reporting. We also contacted trial authors to request missing statistics (SDs/standard errors). However, when authors proved no further missing statistical data, we calculated missing statistics based on methods outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2022a). We did not impute missing outcome data as we consider this could introduce serious bias. The RoB 2 assessment considered missing data due to attrition (Assessment of risk of bias in included studies), and we examined its impact in the Discussion.

Assessment of heterogeneity

We used the I² statistic to quantify inconsistency and measure statistical heterogeneity among the trials but acknowledge that there is substantial uncertainty in the value of I² statistic when there is only a few studies. We also considered the P value from the Chi² test. We were unable to explore possible causes of heterogeneity using subgroup analysis as we had too few trials. Approximate thresholds for interpretation of the I² statistic outlined in the Cochrane Handbook for Systematic Reviews of Interventions are: 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%: considerable heterogeneity (Deeks 2022). Due to few included studies, the contribution of each study to the statistical heterogeneity was not explored visually using the graphical display of study heterogeneity plots neither was the leave‐one‐out sensitivity analysis performed by literally removing one study at a time and recalculating the pooled effect size.

Assessment of reporting biases

We attempted to minimize reporting bias by searching for and including published and unpublished trials. We searched for unpublished trials from grey literature searches, conference abstracts, and trial registries. We were unable to pool more than 10 trials, thus possible publication biases for the primary outcomes through funnel plots or Egger's test was not analyzed.

Data synthesis

We undertook meta‐analyses only where they were meaningful (i.e. if the treatment/supplementation, participants, and underlying clinical questions were similar enough for pooling to make sense). We uploaded the outcome data for each study into the data tables in Review Manager to calculate the pooled effect size using a random‐effects model as there was substantial heterogeneity among studies (RevMan 2024).

We expressed uncertainty by reporting 95% CIs for all effect estimates. Means and SDs were available for all included trials. We also considered the possibility and implications of skewed data when analyzing continuous outcomes, as they can mislead results due to a small sample size (Piovesana 2018). We reported the overall pooled estimate.

For outcomes where it was not deemed possible to perform meta‐analyses, we summarized trial findings narratively.

All eligible trials were included in the primary analysis regardless of the risk of bias.

Subgroup analysis and investigation of heterogeneity

We were unable to investigate whether various study characteristics could explain the clinical diversity or heterogeneity since the number of studies included did not meet the threshold of 10 for meta‐regression. For future updates, we plan to pursue the below subgroup analyses and meta‐regressions as predefined in the published protocol (Campisi 2021b).

• Type of omega‐3 PUFA (to explore whether effects due to the composition of the supplement may alter treatment outcomes and recommendations)

• Adjunctive therapy

• Age of participants

• Duration of omega‐3 fatty acid supplementation

• Duration of follow‐up

• Percentage of female participants

We will conduct subgroup analysis and meta‐regression in R version 4.0.2 using the 'met', 'metafor', and 'dmetar' packages (R).

Sensitivity analysis

We carried out the following sensitivity analyses to test whether key methodologic factors or decisions affected the main results of the primary outcomes.

• To exclude studies with high dropout rates (more than 20%) and compare this against the overall analysis, to determine the extent to which attrition rates might have affected the main result.

• To exclude studies at a high risk of bias and compare this against the main result, to determine the impact of study limitations on the overall effect estimate.

Summary of findings and assessment of the certainty of the evidence

We prepared a Summary of findings table that includes the following outcomes, measured at the end of each trial.

• Depression symptoms, self‐reported (continuous)

• Depression remission (dichotomous)

• Attrition (dichotomous)

• Adverse effects (dichotomous)

Three review authors (AS, CZ, SCC) independently judged the certainty of the evidence, and resolved disagreements by discussion or involving a third review author (DJK).

We used the five GRADE considerations (study limitations based on RoB 2 judgments, inconsistency of effect, imprecision, indirectness, and publication bias) to assess the certainty of the body of evidence as it related to the studies which contributed data to the meta‐analyses for the above outcomes. We used methods and recommendations described in the Cochrane Handbook for Systematic Reviews of Interventions (Schünemann 2022), and developed the table using GRADEpro GDT. We justified all decisions to downgrade the certainty of evidence using footnotes, and we made comments to aid the reader's understanding of the review where necessary.

Results

Description of studies

A detailed description of studies can be found in the Characteristics of included studies; Characteristics of excluded studies; and Characteristics of ongoing studies tables.

Results of the search

The search identified 5783 potential records from the databases. After duplicates were removed, 4454 records remained. Title and abstract screening resulted in the removal of 4378 records. We retrieved the full‐text articles for 77 potentially eligible records, and included 23 relevant records reporting five included studies (Characteristics of included studies table). We excluded 48 records. Reasons for exclusion are outlined in the Characteristics of excluded studies table. We identified four ongoing trials (six records; Characteristics of ongoing studies table). Full details of the study selection process are outlined in the PRISMA flow diagram (Figure). Grey literature searches did not yield any additional records.

Included studies

Characteristics of the five included trials are outlined in the Characteristics of included studies table.

Study design

All studies were parallel‐group RCTs.

Sample size

The five studies involved 228 participants. All trials had small sample sizes: 28 in Nemets 2006; 36 in Fristad 2019; 48 in Gabbay 2018; 56 in McNamara 2020, and 60 in Trebaticka 2020.

Settings

The studies were conducted in high‐income countries. More specifically, three trials were conducted in the USA (Fristad 2019; Gabbay 2018; McNamara 2020), one in Israel (Nemets 2006), and one in Slovakia (Trebaticka 2020).

Three trials recruited participants from hospital or clinical settings (McNamara 2020; Nemets 2006; Trebaticka 2020); one trial recruited from community and clinic settings (Fristad 2019); and one trial did not specify recruitment setting (Gabbay 2018).

Participants

Three trials included children and adolescents with the mean age of 11.6 (SD 2.1) years (Fristad 2019), 14.1 (SD 3.0) years (McNamara 2020), and 10.1 years (Nemets 2006). Two trials included only adolescents with mean ages of 15.6 (SD 1.6) years (Trebaticka 2020), and 16.1 (SD 2.1) years (Gabbay 2018).

There were considerable differences in sex distribution in the included trials. One trial was predominately female: 80% female (McNamara 2020). Two trials were closely split between male and female: 43% female (Fristad 2019) and 57% female (Gabbay 2018). One trial was predominately male: 25% female (Nemets 2006). One trial did not report the sex distribution of participants (Trebaticka 2020).

All trials included participants with common comorbid disorders: Fristad 2019 included those with anxiety, attention deficit hyperactivity disorder, and disruptive behavior disorders; Trebaticka 2020 included those with anxiety; Nemets 2006 included those with anxiety, attention deficit hyperactivity disorder, dysthymia, or tic syndrome; Gabbay 2018 and McNamara 2020 included those with attention deficit hyperactivity disorder.

Trebaticka 2020 allowed people to participate if they used adjunctive antidepressant medication, which may have impacted depression symptoms. The other trials did not take adjunctive antidepressants (Fristad 2019; Gabbay 2018; McNamara 2020; Nemets 2006). All trials allowed participants to continue a course of cognitive behavioral therapy.

Interventions

Omega‐3 PUFA interventions were EPA–DHA combinations and varied considerably among trials. In Fristad 2019, the total daily dose was omega‐3 PUFA 1870 mg (two 500 mg capsules, each containing EPA 350 mg plus DHA 50 mg plus 100 other omega‐3 PUFA) for an EPA:DHA ratio of 7:1. In Gabbay 2018, the total daily dose maintained a 2:1 EPA:DHA ratio that was titrated by 0.6 g/day every two weeks, starting at 1.2 g/day to a maximum of 3.6 g/day (EPA 2.4 g plus DHA 1.2 g). In McNamara 2020, the total daily dose was 2130 mg (EPA 450 mg plus DPA 40 mg plus DHA 260 mg) for an EPA:DHA ratio of 1.73:1. In Nemets 2006, the daily dose was two 500 mg capsules (EPA 190 mg plus DHA 90 mg) or one 1000 mg capsule (EPA 400 mg plus DHA 200 mg) for an EPA:DHA ratio of 2:1. In Trebaticka 2020, the daily volume of emulsion was 20 mL (giving a dose of EPA 1000 mg plus DHA 750 mg) for an EPA:DHA ratio of 1.33:1.

Comparators

Gabbay 2018 used a 50:50 ratio of corn and soybean oils, omega‐6 (50%), monounsaturated (25%) fatty acids as placebo; in McNamara 2020, each placebo capsule contained linoleic acid 92 mg, stearic acid 23 mg, palmitic acid 114 mg, and oleic acid 746 mg; in Nemets 2006, placebo for the 500 mg capsule was olive oil and for the 1000 mg capsule was safflower oil; Trebaticka 2020 used an identical‐looking comparator of sunflower oil emulsion containing omega‐6 linoleic acid 2467 mg. Fristad 2019 did not report the contents of the placebo capsules but stated "without any active treatment" and that they were matched to the omega‐3 intervention for odor and appearance.

Trial duration was 10 weeks (Gabbay 2018), 12 weeks (Fristad 2019; McNamara 2020; Trebaticka 2020), and 16 weeks (Nemets 2006).

Excluded studies

We excluded 48 full‐text records (see Characteristics of excluded studies table). Twenty‐six studies included adults, 15 did not measure depression symptoms, and seven were not RCTs.

Studies awaiting classification

No trials are currently awaiting classification.

Ongoing trials

Our search identified four ongoing trials (see Characteristics of ongoing studies table). The primary author of Haberling 2019 indicated in July 2023 that the analysis was complete. Results from Haberling 2019 will be relevant to future updates as their sample size is the largest (about 300) and conducted among children and adolescents to date. Results from Rice 2016 will be relevant to future updates but have not been published and correspondence in May 2021 did not yield further details. Two additional studies are currently recruiting participants (Cao 2021; Li 2022).

Risk of bias in included studies

Risk of bias assessment for our primary and secondary outcomes at the end of each trial (depression symptomology (continuous), resolution of depressive disorder/remission (dichotomous), attrition (dichotomous)) was applied and differed between outcomes (see below). Furthermore, the overall risk of bias varied between trials, ranging from some concerns to high risk. We have shared the data from the Excel RoB 2 spreadsheet on Figshare (Campisi 2023).

Bias arising from the randomization process

All trials indicated that participants were randomized to either the intervention or control groups. Random sequence generation was at low risk of bias in four trials. Fristad 2019 and Trebaticka 2020 used block randomization. Gabbay 2018 and McNamara 2020 used stratified randomization methods. One trial did not give any information about the allocation method aside from the fact that it was randomized, thus, this trial was judged to have some concerns (Nemets 2006).

Bias due to deviations from intended interventions

All trials were at low risk of bias due to deviations from intended interventions. In detail, all participants as well as study team members were unaware of assigned interventions, and trials used appropriate analyses to estimate the effect of assignment to intervention.

Bias due to missing outcome data

One trial had adequately complete outcome data (3% attrition) and was at low risk of bias (Trebaticka 2020), while another trial had incomplete data but was judged as having some concerns of risk of bias due to its relatively moderate attrition of 19% (Gabbay 2018). The remaining three trials had a high risk of bias due to missing outcome data: McNamara 2020 had 25% attrition; Fristad 2019 had 28% attrition; and Nemets 2006 had 29% attrition.

Bias in measurement of the outcome

All trials used validated measures for the assessment of depression symptoms in intervention and comparator groups, ensuring that all study groups were treated equally, and that assessors were not aware of the treatment assignment. Therefore, all trials had a low risk of bias in assessing depression symptoms.

Bias in selection of the reported result

All trials were judged to have some concerns of risk of bias in the selection of the reported result due to failure in prespecifying the analysis plan before unblinded outcome data were available. However, all trials ensured that the results were reported adequately.

Overall risk of bias

Two trials were judged to have some concerns about overall bias: Trebaticka 2020 had low attrition of 3% but failed to use a prespecified analysis plan, while Gabbay 2018 had moderate attrition of 19% and no pre‐existing analysis plan. However, the remaining three studies exhibited an overall high risk of bias: Fristad 2019 had 28% attrition and no pre‐existing analysis plan; McNamara 2020 exhibited 25% attrition and no pre‐existing analysis plan; Nemets 2006 did not inform about concealment of allocation sequence, had 29% attrition, and no predefined analysis plan.

Effects of interventions

See: Summary of findings 1 Summary of findings table ‐ Omega‐3 polyunsaturated fatty acid (PUFA) compared to placebo for depression in children and adolescents

Primary outcomes

Self‐reported depression symptoms (continuous measure)

All studies reported depression symptoms using continuous data. All trials reported data at baseline and study end. Four studies used the clinician‐rated CDRS‐R (Fristad 2019; Gabbay 2018; McNamara 2020; Nemets 2006), with Gabbay 2018 and Nemets 2006 using additional self‐report scales to measure depression symptomology: the CDI and the BDI‐II. For our meta‐analysis, we used findings measured by the CDRS‐R from Gabbay 2018 and Nemets 2006 to be consistent with the majority of trials and decrease heterogeneity due to depression measures. Trebaticka 2020 used the CDI but was included in the meta‐analysis after ensuring comparable direction. Most trials reported depression symptom scores at multiple time points.

Depression symptoms were measured at baseline and the end of the trial. Omega‐3 PUFA supplementation may reduce self‐reported depression symptoms when compared to control, but the evidence is very uncertain (SMD −0.34, 95% CI −0.85 to 0.17; I² = 63%; 5 RCTs, 185 participants; very low‐certainty evidence; Analysis 1.1). An SMD of 0.2 represents a small effect, 0.5 a moderate effect, and 0.8 a large effect. Between‐trial heterogeneity variance was substantial, meaning that over half of the variation was due to between‐trial heterogeneity. See Table.

1.1. Analysis.

Comparison 1: Primary outcomes, Outcome 1: Self‐reported depression symptoms (continuous measure)

Depression remission (dichotomous measure)

Four trials reported depression as a dichotomous measure (Fristad 2019; Gabbay 2018; McNamara 2020; Nemets 2006). All four trials utilized the same or similar cutoff score on the CDRS‐R to determine remission from depression. Both McNamara 2020 and Fristad 2019 used a CDRS‐R score of 28 or less as the cutoff for indicating remission from depression symptoms. Nemets 2006 employed a similar cutoff, specifying a CDRS score less than 29 as the remission criterion. In Gabbay 2018, treatment responders (i.e. improvement in depression) were based on two criteria: a 50% or more decrease from the baseline CDRS‐R score (after subtracting the 17‐item base score), or a CDRS‐R score of 28 or less at the final visit. This consistency in cutoff scores across these studies suggests a shared approach to defining remission in depression using the CDRS‐R.

Omega‐3 PUFA supplementation may have little to no effect on remission of depression symptoms compared to placebo, but the evidence is very uncertain (OR 1.11, 95% CI 0.45 to 2.75; I² = 26%; 4 RCTs, 127 participants; very low‐certainty evidence; Analysis 1.2). The between‐trial heterogeneity variance was moderate. In Fristad 2019, 45% of participants in the omega‐3 PUFA group achieved remission status compared to a 53% remission rate in the control group. In Gabbay 2018, 9/18 (50%) participants in the omega‐3 PUFA group and 13/21 (62%) participants in the control group were classified as treatment responders. In McNamara 2020, there was remission from depression symptoms by 62% of participants in the fish oil group and 52% of participants in the control group. In Nemets 2006, 4/10 (40%) children in the omega‐3 group met the remission criteria of a CDRS‐R score less than 29 at study exit, while 0/10 children in the control group met this criterion.

1.2. Analysis.

Comparison 1: Primary outcomes, Outcome 2: Depression remission (dichotomous measure)

Secondary outcomes

Attrition (dropouts)

All five studies reported attrition or failure to complete.

Omega‐3 PUFA supplementation may result in little to no difference in attrition (dropouts) (OR 0.94, 95% CI 0.46 to 1.90; I² = 0%; 5 RCTs, 228 participants; low‐certainty evidence; Analysis 2.1). There was no between‐study heterogeneity. One trial had low levels of attrition: Trebaticka 2020 lost 1/30 participants from each group, accounting for 3% of overall attrition. Another trial had moderate levels of attrition: Gabbay 2018 lost 3/21 participants from the omega‐3 and 6/27 participants from the control group, resulting in an overall attrition of 19%. The remaining three trials had high levels of attrition: McNamara 2020 lost 6/27 participants from the fish oil group and 8/29 participants from the control group (overall attrition: 25%); Fristad 2019 lost 7/18 participants from the omega‐3 group and 3/18 participants from the control group (overall attrition: 28%); and Nemets 2006 lost 3/13 participants from the omega‐3 group and 5/15 participants from the control group (overall attrition: 29%). The cause of attrition was not reported to be associated with any adverse effects or dosage in any of the included trials.

2.1. Analysis.

Comparison 2: Secondary outcomes, Outcome 1: Attrition (dropouts)

Adverse effects

All five studies reported no serious adverse effects among children and adolescents taking omega‐3 PUFA supplements. Data regarding other adverse effects are reported below as a narrative summary.

Fristad 2019 found adverse effects to be either absent or mild as measured by a scale of severity (0 [absent] to 6 [severe]), and there were no differences between groups for any of the eight monitored adverse effects (constipation, diarrhea, stomach ache, increased appetite, decreased appetite, burping, fishy breath, or nausea). McNamara 2020 monitored adverse effects for the fish oil and placebo groups with a difference reported for muscle cramps, which were more frequently reported in the fish oil (13/27 participants) than in the placebo group (6/29) participants (P = 0.03). Trebaticka 2020 reported 1/29 participants from the omega‐3 group experienced more frequent defecation (two or three times daily). In the placebo group of Gabbay 2018, 1/27 participants reported mild skin rash, and 1/27 participants developed unusual manic behavior. Nemets 2006 reported no "clinically relevant" adverse effects.

Compliance

Four studies reported compliance (Fristad 2019; Gabbay 2018; McNamara 2020; Trebaticka 2020). However, data were not reported in a usable format.

Trebaticka 2020 assessed compliance by monitoring the volume of emulsion returned, which was within the expected volume amounts, exceeding 95% compliance. However, the specific rate of compliance was not reported. Fristad 2019 measured compliance through pill counts from returned pill boxes at each assessment. The compliance rate ranged from 48% to 100% (mean 86% [SD 15%]). There was no difference between groups. Gabbay 2018 performed capsule counts and recorded them at each visit. Compliance rate for the entire study population ranged from 0.58 to 1.00, with an overall mean compliance rate of 0.89 (SD 0.11). There was no difference in compliance between the two groups (P = 0.21). McNamara 2020 monitored compliance using capsule counts (low‐dose) or bottle volumes (high‐dose) at weekly visits. There was a compliance rate of 93% for the fish oil group and 97% for the placebo group.

Subgroup analyses and meta‐regression

Methods for conducting planned subgroup analysis and meta‐regressions are outlined in the Differences between protocol and review section. For the current review, we were unable to conduct any subgroup analyses or meta‐regression analyses as we included fewer than 10 trials, and the number of trial participants was small.

Sensitivity analyses

We conducted two sensitivity analyses to examine the impact of attrition (dropouts) and risk of bias on the assessment of self‐reported depression symptoms.

Three studies reported high attrition rates (i.e. greater than 20%) (Fristad 2019; McNamara 2020; Nemets 2006). The sensitivity analysis by excluding these three studies showed the effect size was no longer similar to the overall analysis and no heterogeneity (SMD 0, 95% CI −0.40 to 0.40; I² = 0%; 2 RCTs, 97 participants). This indicates the results were significantly influenced by the studies with high attrition.

For each risk‐of‐bias domain, we systematically removed studies with a high risk of bias, leaving only trials with some concerns and low risk of bias. We removed three studies from this sensitivity analysis due to the high risk of bias for missing outcome data and overall bias (Fristad 2019; McNamara 2020; Nemets 2006). The sensitivity analysis excluding these three studies showed the effect size was no longer similar to the overall analysis (SMD 0, 95% CI −0.40 to 0.40; I² = 0%; 2 RCTs, 97 participants). This indicates the results were significantly influenced by the studies with a high risk of bias.

We were unable to perform a sensitivity analysis on depression remission (as a dichotomous measure) since removing the studies with a high risk of bias or high attrition left only one study.

Discussion

Summary of main results

The aim of the current review was to determine the benefits and harms of omega‐3 PUFA in reducing depression symptoms among children and adolescents with clinical depression. Overall, this review highlights an absence of evidence for omega‐3 PUFA supplementation in the treatment of depression for children and adolescents, with only five trials and small sample sizes (a total of 228 participants). See Table. Omega‐3 PUFA may reduce self‐reported depression symptoms compared to placebo, but the evidence is very uncertain. Omega‐3 PUFA may have little to no effect on depression remission as a dichotomous measure, but the evidence is very uncertain. The evidence suggests that omega‐3 PUFA may result in little to no difference in attrition (dropouts) or adverse effects. Data on compliance were limited.

Overall completeness and applicability of evidence

To date, few RCTs have investigated the effectiveness of omega‐3 PUFA in children and adolescents with depression. The trials included in the meta‐analysis used similar measures of depression symptomology (i.e. four of the five used CDRS‐R), were of similar duration (i.e. greater than 10 weeks), had small sample sizes, and had similar rates of attrition and few adverse effects. There was variability in the omega‐3‐PUFA dosages (i.e. EPA:DHA ratios, and different delivery modes, i.e. capsule, emulsion).

A particular limitation was that all trials had very small sample sizes (60 participants or fewer per trial). A major contributing factor to the lack of observed effect may be the result of insufficient power due to the small sample size. Another plausible factor for the lack of observed effect may be the relatively short duration of the trials. Several pathways have been hypothesized for omega‐3 PUFA's impact on depression symptomology. The minimum supplementation duration remains largely unknown, except for omega‐3 PUFA equilibration in the brain, which requires a minimum six‐month duration of supplementation (Browning 2012).

Sensitivity analyses after considering dropouts and risk of bias in included studies differed from the full analysis. Excluding studies that had high attrition rates or high risk of bias resulted in an SMD of 0 (95% CI −0.40 to 0.40), suggesting that the results were significantly influenced by studies with a high risk of bias or high levels of attrition.

We were unable to complete any subgroup analyses or meta‐regression since the number of included trials was less than the threshold of 10 studies.

Quality of the evidence

The certainty of the evidence for all outcome measures was low or very low. Our judgments of certainty according to GRADE are given in Table.

The certainty of evidence for our primary outcomes, depression symptoms (as a continuous measure) and remission rates, was very low. This very‐low level of certainty resulted primarily from the high risk of bias and imprecision, as the width of the CIs encompassed both potentially significant benefits and possible harm. Moreover, the body of evidence consisted mainly of limited and predominantly small studies, where there was significant heterogeneity and remained unexplained by any other analysis. Regarding the secondary outcome, attrition, the certainty of evidence was low due to imprecision. There were no concerns regarding the risk of bias, inconsistency, or indirectness. However, the wide CIs indicated a broad range of possible effects.

For all outcomes, funnel plots could not be generated and thus publication bias could not be assessed.

Potential biases in the review process

Overall, the risk of bias in the trials was some concerns or high. In detail, all studies failed to include a prespecified analysis plan that was finalized before outcome data became available. Trebaticka 2020 and Gabbay 2018 had low and moderate attrition, respectively, and were therefore rated to have some concerns of bias. However, Fristad 2019, McNamara 2020, and Nemets 2006 showed dropout rates of more than 20%, leading to high risk of bias for all three trials. Additionally, Nemets 2006 failed to confirm that the allocation sequence was concealed until participants were enrolled and assigned to the different treatment groups.

The findings in this review may have been influenced by bias due to the limited number of trials included and the small sample sizes. It is worth noting that all the included studies shared similarly small sample sizes, and as a result, the relative weighting in all analyses ranged from 14.9% to 23.2%. Consequently, the results of the analyses were not significantly influenced by any individual study. However, it is important to interpret the findings with caution, considering the overall small number of trials and their sample sizes.

While our search was not limited to English‐language articles, it is important to acknowledge the inherent bias in electronic database searches, which tend to favor articles published in English and mainstream journals. This bias is a common issue in all types of reviews, including this one. To mitigate this bias, we attempted to broaden our search beyond English language articles. However, despite these efforts, we were unable to find any non‐English studies that met the inclusion criteria.

Agreements and disagreements with other studies or reviews

To date, two reviews have been conducted specifically on children and adolescents examining the impact of omega‐3 PUFA on depression symptoms (Lopresti 2015; Zhang 2019). Zhang 2019 found that omega‐3 PUFA was no better than placebo on end‐point depression symptoms (SMD −0.12, 95% CI −0.53 to 0.30; P = 0.58). In our view and as mentioned in Why it is important to do this review, the eligibility criteria applied in Zhang 2019 may have been overly restrictive as it only included studies with placebo as the comparator and excluded participants who had comorbid physical, psychiatric, or combinations of physical and psychiatric diseases, whereas our current review considered all comparators and participants with other psychiatric diseases. Lopresti 2015 is a narrative review citing two trials (McNamara 2020; Nemets 2006), thus providing limited, systematically synthesized evidence to evaluate the benefits and harms of omega‐3 PUFA, but Lopresti 2015 noted small sample sizes as significant limitations of their review.

Authors' conclusions

Implications for practice.

This review highlights the nascent nature of research relating to omega‐3 PUFA and depression in children and adolescents. The currently available evidence is very uncertain about the effect of omega‐3 PUFA supplementation in improving self‐reported depression symptoms or depression remission among children and adolescents. Although current evidence shows that omega‐3 PUFA supplementation may result in little to no difference in adverse effects, the evidence is of low certainty and thus there is insufficient information to draw definitive conclusions regarding potential harms.

Implications for research.

Much remains to be determined regarding the potential impact of omega‐3 PUFA on depression symptoms in children and adolescents. Future high‐quality study designs must address several limitations in the current body of literature including larger sample sizes, dosage, eicosapentaenoic acid/docosahexaenoic acid ratios (omega‐3 index), treatment durations, and safety/adverse effects.

History

Protocol first published: Issue 6, 2021

Risk of bias

Risk of bias for analysis 1.1 Self‐reported depression symptoms (continuous measure).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Fristad 2019

Low risk of bias

Allocation sequence was random and concealed; no baseline differences between intervention groups.

Low risk of bias

Intervention assignment was unknown by participants and study staff; appropriate analysis was used.

High risk of bias

Attrition was at 28%; missingness in the outcome could depend on its true value.

Low risk of bias

Outcome measurement was appropriate and did not differ between groups; outcome assessors were unaware of treatment assignment.

Some concerns

Outcome data were selected from one outcome measure using one analysis; outcome data were not analysed according to a pre‐specified plan.

High risk of bias

The study had 28% attrition and no pre‐specified analysis plan, resulting in high risk of overall bias for depressive symptoms.

Gabbay 2018

Low risk of bias

Allocation sequence was random and concealed; no baseline differences between intervention groups.

Low risk of bias

Intervention assignment was unknown by participants and study staff; appropriate analysis was used.

Some concerns

Attrition was at 19%; missingness in the outcome could depend on its true value.

Low risk of bias

Outcome measurement was appropriate and did not differ between groups; outcome assessors were unaware of treatment assignment.

Some concerns

Outcome data were selected from one outcome measure using one analysis; outcome data were not analysed according to a pre‐specified plan.

Some concerns

The study had 19% attrition and no pre‐specified analysis plan, leading to some concerns of risk of overall bias for depressive symptoms.

McNamara 2020

Low risk of bias

Allocation sequence was random and concealed; no baseline differences between intervention groups.

Low risk of bias

Intervention assignment was unknown by participants and study staff; appropriate analysis was used.

High risk of bias

Attrition was at 25%; missingness in the outcome could depend on its true value.

Low risk of bias

Outcome measurement was appropriate and did not differ between groups; outcome assessors were unaware of treatment assignment.

Some concerns

Outcome data were selected from one outcome measure using one analysis; outcome data were not analysed according to a pre‐specified plan.

High risk of bias

The study had 25% attrition and no pre‐specified analysis plan, resulting in high risk of overall bias for depressive symptoms.

Nemets 2006

Some concerns

Allocation sequence was random but no information on concealment of the allocation sequence; no baseline differences between intervention groups.

Low risk of bias

Intervention assignment was unknown by participants and study staff; appropriate analysis was used.

High risk of bias

Attrition was at 29%; missingness in the outcome could depend on its true value.

Low risk of bias

Outcome measurement was appropriate and did not differ between groups; outcome assessors were unaware of treatment assignment.

Some concerns

Outcome data were selected from one outcome measure using one analysis; outcome data were not analysed according to a pre‐specified plan.

High risk of bias

The study had 29% attrition, no pre‐specified analysis plan, and no information if the allocation sequence was concealed until participants were enrolled and assigned to interventions, resulting in high risk of overall bias for depressive symptoms.

Trebaticka 2020

Low risk of bias

Allocation sequence was random and concealed; no baseline differences between intervention groups.

Low risk of bias

Intervention assignment was unknown by participants and study staff; appropriate analysis was used.

Low risk of bias

Outcome data for almost all participants were reported (3% attrition).

Low risk of bias

Outcome measurement was appropriate and did not differ between groups; outcome assessors were unaware of treatment assignment.

Some concerns

Outcome data were selected from one outcome measure using one analysis; outcome data were not analysed according to a pre‐specified plan.

Some concerns

The study had no pre‐defined analysis plan, leading to some concerns of risk of overall bias for depressive symptoms.

Risk of bias for analysis 1.2 Depression remission (dichotomous measure).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Fristad 2019

Low risk of bias

Allocation sequence was random and concealed; no baseline differences between intervention groups.

Low risk of bias

Intervention assignment was unknown by participants and study staff; appropriate analysis was used.

High risk of bias

Attrition was at 28%; missingness in the outcome could depend on its true value.

Low risk of bias

Outcome measurement was appropriate and did not differ between groups; outcome assessors were unaware of treatment assignment.

Some concerns

Outcome data were selected from one outcome measure using one analysis; outcome data were not analysed according to a pre‐specified plan.

High risk of bias

The study had 28% attrition and no pre‐specified analysis plan, resulting in high risk of overall bias for depressive symptoms.

Gabbay 2018

Low risk of bias

Allocation sequence was random and concealed; no baseline differences between intervention groups.

Low risk of bias

Intervention assignment was unknown by participants and study staff; appropriate analysis was used.

Some concerns

Attrition was at 19%; missingness in the outcome could depend on its true value.

Low risk of bias

Outcome measurement was appropriate and did not differ between groups; outcome assessors were unaware of treatment assignment.

Some concerns

Outcome data were selected from one outcome measure using one analysis; outcome data were not analysed according to a pre‐specified plan.

Some concerns

The study had 19% attrition and no pre‐specified analysis plan, leading to some concerns of risk of overall bias for depressive symptoms.

McNamara 2020

Low risk of bias

Allocation sequence was random and concealed; no baseline differences between intervention groups.

Low risk of bias

Intervention assignment was unknown by participants and study staff; appropriate analysis was used.

High risk of bias

Attrition was at 25%; missingness in the outcome could depend on its true value.

Low risk of bias

Outcome measurement was appropriate and did not differ between groups; outcome assessors were unaware of treatment assignment.

Some concerns

Outcome data were selected from one outcome measure using one analysis; outcome data were not analysed according to a pre‐specified plan.

High risk of bias

The study had 25% attrition and no pre‐specified analysis plan, resulting in high risk of overall bias for depressive symptoms.

Nemets 2006

Some concerns

Allocation sequence was random but no information on concealment of the allocation sequence; no baseline differences between intervention groups.

Low risk of bias

Intervention assignment was unknown by participants and study staff; appropriate analysis was used.

High risk of bias

Attrition was at 29%; missingness in the outcome could depend on its true value.

Low risk of bias

Outcome measurement was appropriate and did not differ between groups; outcome assessors were unaware of treatment assignment.

Some concerns

Outcome data were selected from one outcome measure using one analysis; outcome data were not analysed according to a pre‐specified plan.

High risk of bias

The study had 29% attrition, no pre‐specified analysis plan, and no information if the allocation sequence was concealed until participants were enrolled and assigned to interventions, resulting in high risk of overall bias for depressive symptoms.

Risk of bias for analysis 2.1 Attrition (dropouts).

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Fristad 2019

Low risk of bias

Allocation sequence was random and concealed; no baseline differences between intervention groups.

Low risk of bias

Intervention assignment was unknown by participants and study staff; appropriate analysis was used.

Low risk of bias

Outcome data for all participants were available.

Low risk of bias

Outcome measurement was appropriate and did not differ between groups; outcome assessors were unaware of treatment assignment.

Some concerns

Outcome data were selected from one outcome measure using one analysis; outcome data were not analysed according to a pre‐specified plan.

Some concerns

The study had no pre‐specified analysis plan, leading to some concerns for risk of overall bias for attrition.

Gabbay 2018

Low risk of bias

Allocation sequence was random and concealed; no baseline differences between intervention groups.

Low risk of bias

Intervention assignment was unknown by participants and study staff; appropriate analysis was used.

Low risk of bias

Outcome data for all participants were available.

Low risk of bias

Outcome measurement was appropriate and did not differ between groups; outcome assessors were unaware of treatment assignment.

Some concerns

Outcome data were selected from one outcome measure using one analysis; outcome data were not analysed according to a pre‐specified plan.

Some concerns

The study had no pre‐specified analysis plan, leading to some concerns for risk of overall bias for attrition.

McNamara 2020

Low risk of bias

Allocation sequence was random and concealed; no baseline differences between intervention groups.

Low risk of bias

Intervention assignment was unknown by participants and study staff; appropriate analysis was used.

Low risk of bias

Outcome data for all participants were available.

Low risk of bias

Outcome measurement was appropriate and did not differ between groups; outcome assessors were unaware of treatment assignment.

Some concerns

Outcome data were selected from one outcome measure using one analysis; outcome data were not analysed according to a pre‐specified plan.

Some concerns

The study had no pre‐specified analysis plan, leading to some concerns for risk of overall bias for attrition.

Nemets 2006

Some concerns

Allocation sequence was random but no information on concealment of the allocation sequence; no baseline differences between intervention groups.

Low risk of bias

Intervention assignment was unknown by participants and study staff; appropriate analysis was used.

Low risk of bias

Outcome data for all participants were available.

Low risk of bias

Outcome measurement was appropriate and did not differ between groups; outcome assessors were unaware of treatment assignment.

Some concerns

Outcome data were selected from one outcome measure using one analysis; outcome data were not analysed according to a pre‐specified plan.

Some concerns

The study had no information on allocation sequence concealment and no pre‐specified analysis plan, leading to some concerns for risk of overall bias for attrition.

Trebaticka 2020

Low risk of bias

Allocation sequence was random and concealed; no baseline differences between intervention groups.

Low risk of bias

Intervention assignment was unknown by participants and study staff; appropriate analysis was used.

Low risk of bias

Outcome data for all participants were available.

Low risk of bias

Outcome measurement was appropriate and did not differ between groups; outcome assessors were unaware of treatment assignment.

Some concerns

Outcome data were selected from one outcome measure using one analysis; outcome data were not analysed according to a pre‐specified plan.

Some concerns

The study had no pre‐specified analysis plan, leading to some concerns for risk of overall bias for attrition.

Appendices

Appendix 1. Diagnostic and Statistical Manual and International Classification of Diseases criteria

When clinical diagnoses are provided, the following Diagnostic and Statistical Manual 5th Edition (DSM‐5) and International Classification of Diseases 11th Edition (ICD‐11) diagnoses are eligible. Comparable diagnoses are acceptable for DSM‐IV and ICD‐10.

DSM‐5 codes and diagnoses

296.36 F33.42 Major depressive disorder, recurrent episode, in full remission

296.35 F33.41 Major depressive disorder, recurrent episode, in partial remission

296.31 F33.0 Major depressive disorder, recurrent episode, mild

296.32 F33.1 Major depressive disorder, recurrent episode, moderate

296.33 F33.2 Major depressive disorder, recurrent episode, severe

296.30 F33.9 Major depressive disorder, recurrent episode, unspecified

296.34 F33.3 Major depressive disorder, recurrent episode, with psychotic features

296.26 F32.5 Major depressive disorder, single episode, in full remission

296.25 F32.4 Major depressive disorder, single episode, in partial remission

296.21 F32.0 Major depressive disorder, single episode, mild

296.22 F32.1 Major depressive disorder, single episode, moderate

296.23 F32.2 Major depressive disorder, single episode, severe

296.20 F32.9 Major depressive disorder, single episode, unspecified

296.24 F32.3 Major depressive disorder, single episode, with psychotic features

300.4 F34.1 Persistent depressive disorder (dysthymia)

311 F32.9 Unspecified depressive disorder

ICD‐11 codes and diagnoses

6A70 Single episode depressive disorder

6A70.0 Single episode depressive disorder, mild

6A70.1 Single episode depressive disorder, moderate, without psychotic symptoms

6A70.2 Single episode depressive disorder, moderate, with psychotic symptoms

6A70.3 Single episode depressive disorder, severe, without psychotic symptoms

6A70.4 Single episode depressive disorder, severe, with psychotic symptoms

6A70.5 Single episode depressive disorder, unspecified severity

6A70.6 Single episode depressive disorder, currently in partial remission

6A70.7 Single episode depressive disorder, currently in full remission

6A70.Y Other specified single episode depressive disorder

6A70.Z Single episode depressive disorder, unspecified

6A71 Recurrent depressive disorder

6A72 Dysthymic disorder

6A73 Mixed depressive and anxiety disorder

6A7Y Other specified depressive disorders

6A7Z Depressive disorders

SD82 Depression disorder

Appendix 2. Search strategies

Ovid MEDLINE: Epub Ahead of Print, In‐Process & Other Non‐Indexed Citations, Ovid MEDLINE Daily and Ovid MEDLINE <1946‐Present>

1 Depression/ 149550

2 Depressive Disorder/ 74977

3 Depressive Disorder, Major/ 38063

4 Dysthymic Disorder/ 1175

5 mental disorders/ 177144

6 Mood Disorders/ 15872

7 Premenstrual Dysphoric Disorder/ 243

8 depress*.tw,kf. 560875

9 dysthymi*.tw,kf. 3326

10 melanchol*.tw,kf. 3296

11 MDD.tw,kf. 18384

12 mood*.tw,kf. 93014

13 (internaliz* or involutional).tw,kf. 71795

14 (affective disorder* or affective symptom*).tw,kf. 20900

15 dysphoric disorder*.tw,kf. 1206

16 (mental adj4 (health or disorder* or illness* or stress or turmoil or anguish or dispair)).tw,kf. 305008

17 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10 or 11 or 12 or 13 or 14 or 15 or 16 1037509

18 pediatrics/ 57830

19 young adult/ 1009043

20 Adolescent/ 2211078

21 Child/ 1904949

22 Child, Preschool/ 988084

23 (minors or boy or boys or boyhood or girl or girls or girlhood or kid or kids or kiddie or child* or schoolchild* or adolescen* or preadolescen* or juvenil* or youth* or teen* or preteen* or underage* or under age* or pubescen* or prepubescen* or puberty or pe?diatric* or p?ediatric* or school*).tw,kf. 2456487

24 (young* adj3 (adult* or person* or individual* or people* or population* or man or men or wom?n)).tw,kf. 288203

25 (youngster* or preschool* or pre‐school* or kindergarten* or freshm?n* or junior* or soph?more* or senior* or highschool* or high school* or colleg* or universit* or student* or undergrad* or grade school* or elementary school* or middle school* or middleschool* or public school* or secondary school* or intermediate school*).tw,kf. 980490

26 ((grade* or form*) adj3 (one or first or two or second or three or third or four* or five or fifth or six* or seven* or eight* or nine* or ninth or ten* or eleven* or twelfth or twelve or "1" or "2" or "3" or "4" or "5" or "6" or "7" or "8" or "9" or "10" or "11" or "12")).tw,kf. 594963

27 18 or 19 or 20 or 21 or 22 or 23 or 24 or 25 or 26 5853466

28 Fatty Acids, Omega‐3/ 15945

29 fish oils/ 8461

30 cod liver oil/ 552

31 docosahexaenoic acids/ 10457

32 eicosapentaenoic acid/ 7185

33 dietary fats, unsaturated/ 5414

34 alpha‐Linolenic Acid/ 3255

35 Dietary Supplements/ 73379

36 (omega3* or omega 3*).tw,kf. 20600

37 fatty acid*.tw,kf. 262306

38 ((n3 or n‐3 or w3 or w‐3) adj3 polyunsaturat*).tw,kf. 6385

39 (PUFA* or n‐3PUFA* or n3PUFA*).tw,kf. 18769

40 ((n3 or n‐3 or w3 or w‐3) adj3 oil*).tw,kf. 937

41 (fish* adj3 oil*).tw,kf. 13033

42 (cod adj3 oil*).tw,kf. 1109

43 icosapentanoic*.tw,kf. 1

44 icosapentaenoic*.tw,kf. 37

45 timnodonic*.tw,kf. 21

46 (1553‐41‐9 or aan7qov9ea).tw,kf. 0

47 (diet* adj3 (fat* or oil*)).tw,kf. 78197

48 (unsaturate* adj3 (fat* or oil*)).tw,kf. 16368

49 (alphalinolen* or alpha‐linolen*).tw,kf. 6452

50 (linolenate* or linolenic*).tw,kf. 13360

51 Epanova.tw,kf. 23

52 ((diet* or food*) adj2 supplement*).tw,kf. 57389

53 (Marine adj3 (oil* or lipid*)).tw,kf. 1763

54 Coromega.tw,kf. 2

55 Efamed.tw,kf. 0

56 MaxEPA.tw,kf. 172

57 Lovaza.tw,kf. 57

58 Omtryg.tw,kf. 2

59 (Vascepa or icosapent ethyl).tw,kf. 317

60 DHA.tw,kf. 17541

61 Docosahex*.tw,kf. 17151

62 (Eicosapent* or ethyl‐eicosapent* or ethyleicosapent*).tw,kf. 12684

63 EPA.tw,kf. 22845

64 (bilantin omega or conchol 36 or eicosa e or eicosapen or sakana).tw,kf. 8

65 28 or 29 or 30 or 31 or 32 or 33 or 34 or 35 or 36 or 37 or 38 or 39 or 40 or 41 or 42 or 43 or 44 or 45 or 46 or 47 or 48 or 49 or 50 or 51 or 52 or 53 or 54 or 55 or 56 or 57 or 58 or 59 or 60 or 61 or 62 or 63 or 64 445598

66 randomized controlled trial.pt. 592754

67 controlled clinical trial.pt. 95302

68 randomi#ed.ab. 720186

69 placebo.ab. 238175

70 drug therapy.fs. 2590624

71 randomly.ab. 408246

72 trial.ab. 648102

73 groups.ab. 2516702

74 clinical trials as topic/ 200972

75 (randomi#ed or randomi#ation or randomi#ing).ti,ab,kf. 801564

76 (RCT or cRCT or "at random").ti,ab,kf. 47297

77 (random* adj3 (administ* or allocat* or assign* or class* or cluster or crossover or cross‐over or control* or determine* or divide* or division or distribut* or expose* or fashion or number* or place* or pragmatic or quasi or recruit* or split or subsitut* or treat*)).ti,ab,kf. 696795

78 placebo.ab,ti,kf. 244974

79 trial.ti. 285130

80 (control* adj3 group*).ab. 658106

81 (control* and (trial or study or group*) and (waitlist* or wait* list* or ((treatment or care) adj2 usual))).ti,ab,kf,hw. 33877

82 ((single or double or triple or treble) adj2 (blind* or mask* or dummy)).ti,ab,kf. 198076

83 double‐blind method/ or random allocation/ or single‐blind method/ 303313

84 66 or 67 or 68 or 69 or 70 or 71 or 72 or 73 or 74 or 75 or 76 or 77 or 78 or 79 or 80 or 81 or 82 or 83 6044128

85 exp animals/ not humans.sh. 5121641

86 84 not 85 5264706

87 17 and 27 and 65 and 86 796

Embase Classic + Embase <1947 to 17 May 2023>

1 depression/ or mood disorder/ or adolescent depression/ or agitated depression/ or atypical depression/ or chronic depression/ or depressive psychosis/ or dysthymia/ or endogenous depression/ or involutional depression/ or major depression/ or melancholia/ or premenstrual dysphoric disorder/ or reactive depression/ or subsyndromal depression/ or treatment resistant depression/ 590668

2 mental disease/ or mood disorder/ 341767

3 depress*.tw,kw. 800081

4 dysthymi*.tw,kw. 4577

5 melanchol*.tw,kw. 4799

6 MDD.tw,kw. 27660

7 mood*.tw,kw. 140547

8 (internaliz* or involutional).tw,kw. 88086

9 (affective disorder* or affective symptom*).tw,kw. 30031

10 (mental adj3 (health or disorder* or illness* or stress or turmoil or anguish or dispair)).tw,kw. 363468

11 dysphoric disorder*.tw,kw. 1538

12 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10 or 11 1521782

13 pediatrics/ 101987

14 young adult/ 507168

15 adolescent/ 1901423

16 child/ or boy/ or girl/ or preschool child/ or school child/ 2697790

17 juvenile/ 64272

18 high school student/ 10186

19 (minors or boy or boys or boyhood or girl or girls or girlhood or kid or kids or kiddie or child* or schoolchild* or adolescen* or preadolescen* or juvenil* or youth* or teen* or preteen* or underage* or under age* or pubescen* or prepubescen* or puberty or pe?diatric* or p?ediatric* or school*).tw,kw. 3367552

20 (young* adj2 (adult* or person* or individual* or people* or population* or man or men or wom?n)).tw,kw. 380522

21 (youngster* or preschool* or pre‐school* or kindergarten* or freshm?n* or junior* or soph?more* or senior* or highschool* or high school* or colleg* or universit* or student* or undergrad* or grade school* or elementary school* or middleschool* or middle school* or public school* or secondary school* or intermediate school*).tw,kw. 1591786

22 ((grade* or form*) adj2 (one or first or two or second or three or third or four* or five or fifth or six* or seven* or eight* or nine* or ninth or ten* or eleven* or twelfth or twelve or "1" or "2" or "3" or "4" or "5" or "6" or "7" or "8" or "9" or "10" or "11" or "12")).tw,kw. 651413

23 school/ or college/ or community college/ or high school/ or kindergarten/ or middle school/ or nursery school/ or primary school/ or university/ 356418

24 13 or 14 or 15 or 16 or 17 or 18 or 19 or 20 or 21 or 22 or 23 6970021

25 omega 3 fatty acid/ 38695

26 fish oil/ 19173

27 cod liver oil/ 1847

28 docosahexaenoic acid/ or unsaturated fatty acid/ 42613

29 icosapentaenoic acid/ 20265

30 linolenic acid/ 12128

31 dietary supplement/ 22406

32 (omega3* or omega 3*).tw,kw. 28707

33 fatty acid*.tw,kw. 324319

34 ((n3 or n‐3 or w3 or w‐3) adj3 polyunsaturat*).tw,kw. 7676

35 (PUFA* or n‐3PUFA* or n3PUFA*).tw,kw. 23346

36 ((n3 or n‐3 or w3 or w‐3) adj3 oil*).tw,kw. 1121

37 (fish* adj2 oil*).tw,kw. 15856

38 (cod adj2 oil*).tw,kw. 1541

39 icosapentanoic*.tw,kw. 2

40 icosapentaenoic*.tw,kw. 78

41 timnodonic*.tw,kw. 20

42 (1553‐41‐9 or aan7qov9ea).tw,kw. 0

43 (diet* adj2 (fat* or oil*)).tw,kw. 97914

44 (unsaturate* adj2 (fat* or oil*)).tw,kw. 18902

45 (alphalinolen* or alpha‐linolen*).tw,kw. 7174

46 (linolenate* or linolenic*).tw,kw. 15206

47 Epanova.tw,kw. 104

48 ((diet* or food*) adj2 supplement*).tw,kw. 70414

49 (Marine adj3 (oil* or lipid*)).tw,kw. 2049

50 (Coromega or Efamed or MaxEPA or Lovaza or Omtryg or Vascepa or icosapent ethyl or DHA or Docosahex* or Eicosapent* or ethyl‐eicosapent* or ethyleicosapent* or EPA or bilatin omega or conchol 36 or eicosa e or eicosapen or sakana).tw,kw. 51234

51 25 or 26 or 27 or 28 or 29 or 30 or 31 or 32 or 33 or 34 or 35 or 36 or 37 or 38 or 39 or 40 or 41 or 42 or 43 or 44 or 45 or 46 or 47 or 48 or 49 or 50 525232

52 Randomized controlled trial/ 787224

53 Controlled clinical study/ 469613

54 random$.ti,ab. 1977979

55 randomization/ 99395

56 intermethod comparison/ 297099

57 placebo.ti,ab. 370913

58 (compare or compared or comparison).ti. 634467

59 ((evaluated or evaluate or evaluating or assessed or assess) and (compare or compared or comparing or comparison)).ab. 2770205

60 (open adj label).ti,ab. 108910

61 ((double or single or doubly or singly) adj (blind or blinded or blindly)).ti,ab. 279857

62 double blind procedure/ 212958

63 parallel group$1.ti,ab. 32206

64 (crossover or cross over).ti,ab. 125813

65 ((assign$ or match or matched or allocation) adj5 (alternate or group$1 or intervention$1 or patient$1 or subject$1 or participant$1)).ti,ab. 416819

66 (assigned or allocated).ti,ab. 491171

67 (controlled adj7 (study or design or trial)).ti,ab. 454053

68 (volunteer or volunteers).ti,ab. 288406

69 human experiment/ 650775

70 trial.ti. 410680

71 or/1‐19 6309640

72 Animal experiment/ not (human experiment/ or human/) 2579732

73 71 not 72 6166321

74 12 and 24 and 51 and 73 2130

APA PsycInfo <1806 to May week 2 2023>

1 "depression (emotion)"/ or major depression/ or sadness/ 175775

2 affective disorders/ or premenstrual dysphoric disorder/ 16342

3 dysthymic disorder/ 1523

4 major depression/ or reactive depression/ or recurrent depression/ or treatment resistant depression/ or atypical depression/ 150803

5 depress*.tw. 356762

6 dysthymi*.tw. 4031

7 melanchol*.tw. 5142

8 MDD.tw. 13627

9 mood*.tw. 86781

10 (internaliz* or involutional).tw. 33795

11 (affective disorder* or affective symptom*).tw. 21431

12 (mental adj3 (health or disorder* or illness* or stress or turmoil or anguish or dispair)).tw. 326283

13 dysphoric disorder*.tw. 969

14 or/1‐13 694512

15 pediatrics/ 31003

16 adolescent development/ 53943

17 adolescent characteristics/ 5254

18 exp childhood development/ 123899

19 developmental age groups/ 8070

20 college students/ or elementary school students/ or high school students/ or junior high school students/ or kindergarten students/ or middle school students/ or preschool students/ 197863

21 (minors or boy or boys or boyhood or girl or girls or girlhood or kid or kids or kiddie or child* or schoolchild* or adolescen* or preadolescen* or juvenil* or youth* or teen* or preteen* or underage* or under age* or pubescen* or prepubescen* or puberty or pe?diatric* or p?ediatric* or school*).tw. 1288228

22 (young* adj3 (adult* or person* or individual* or people* or population* or man or men or wom?n)).tw. 137971

23 (youngster* or preschool* or pre‐school* or kindergarten* or freshm?n* or junior* or soph?more* or senior* or highschool* or high school* or colleg* or universit* or student* or undergrad* or grade school* or elementary school* or middle school* or middleschool* or public school* or secondary school* or intermediate school*).tw. 924966

24 ((grade* or form*) adj3 (one or first or two or second or three or third or four* or five or fifth or six* or seven* or eight* or nine* or ninth or ten* or eleven* or twelfth or twelve or "1" or "2" or "3" or "4" or "5" or "6" or "7" or "8" or "9" or "10" or "11" or "12")).tw. 140348

25 or/15‐24 1926751

26 fatty acids/ 3199

27 dietary supplements/ 2203

28 (omega3* or omega 3*).tw. 1631

29 fatty acid*.tw. 5316

30 ((n3 or n‐3 or w3 or w‐3) adj3 polyunsaturat*).tw. 295

31 (PUFA* or n‐3PUFA* or n3PUFA*).tw. 756

32 ((n3 or n‐3 or w3 or w‐3) adj3 oil*).tw. 11

33 (fish* adj3 oil*).tw. 362

34 (cod adj3 oil*).tw. 28

35 icosapentanoic*.tw. 0

36 icosapentaenoic*.tw. 0

37 timnodonic*.tw. 0

38 (1553‐41‐9 or aan7qov9ea).tw. 0

39 (diet* adj3 (fat* or oil*)).tw. 3596

40 (unsaturate* adj3 (fat* or oil*)).tw. 203

41 (alphalinolen* or alpha‐linolen*).tw. 131

42 (linolenate* or linolenic*).tw. 198

43 Epanova.tw. 1

44 ((diet* or food*) adj2 supplement*).tw. 2313

45 (Marine adj3 (oil* or lipid*)).tw. 10

46 Coromega.tw. 0

47 Efamed.tw. 0

48 MaxEPA.tw. 1

49 Lovaza.tw. 4

50 Omtryg.tw. 0

51 (Vascepa or icosapent ethyl).tw. 2

52 DHA.tw. 842

53 Docosahex*.tw. 917

54 (Eicosapent* or ethyl‐eicosapent* or ethyleicosapent*).tw. 525

55 EPA.tw. 927

56 (bilantin omega or conchol 36 or eicosa e or eicosapen or sakana).tw. 4

57 or/26‐56 13009

58 14 and 25 and 57 456

CENTRAL

ID Search Hits

#1 [mh ^"depression"] 18299

#2 [mh ^"Depressive Disorder"] 9123

#3 [mh ^"depressive disorder, major"] 6441

#4 [mh ^"Dysthymic Disorder"] 195

#5 [mh ^"mental disorders"] 4913

#6 [mh ^"Mood Disorders"] 1015

#7 [mh ^"Premenstrual Dysphoric Disorder"] 50

#8 depress*:ti,ab,kw 106921

#9 dysthymi*:ti,ab,kw 1038

#10 melanchol*:ti,ab,kw 390

#11 MDD:ti,ab,kw 4496

#12 mood*:ti,ab,kw 25345

#13 (internaliz* or involutional):ti,ab,kw 1795

#14 (affective disorder* or affective symptom*):ti,ab,kw 5185

#15 dysphoric disorder*:ti,ab,kw 462

#16 (mental NEAR/2 (health or disorder* or illness* or stress or turmoil or anguish or dispair)):ti,ab,kw 43426

#17 {OR #1‐#16} 149163

#18 [mh ^"pediatrics"] 1070

#19 [mh ^"young adult"] 84591

#20 [mh ^"Adolescent"] 125309

#21 [mh ^"Child"] 69053

#22 [mh ^"Child, Preschool"] 35052

#23 (minors or boy or boys or boyhood or girl or girls or girlhood or kid or kids or kiddie or child* or schoolchild* or adolescen* or preadolescen* or juvenil* or youth* or teen* or preteen* or underage* or under age* or pubescen* or prepubescen* or puberty or pe?diatric* or p?ediatric* or school*):ti,ab,kw 388203

#24 (young* NEAR/2 (adult* or person* or individual* or people* or population* or man or men or wom?n)):ti,ab,kw 114425

#25 (youngster* or preschool* or pre‐school* or kindergarten* or freshm?n* or junior* or soph?more* or senior* or highschool* or high school* or colleg* or universit* or student* or undergrad* or grade school* or elementary school* or middle school* or middleschool* or public school* or secondary school* or intermediate school*):ti,ab,kw 179029

#26 ((grade* or form*) NEAR/2 (one or first or two or second or three or third or four* or five or fifth or six* or seven* or eight* or nine* or ninth or ten* or eleven* or twelfth or twelve or "1" or "2" or "3" or "4" or "5" or "6" or "7" or "8" or "9" or "10" or "11" or "12")):ti,ab,kw 75798

#27 {OR #18 ‐ #26} 1046392

#28 [mh ^"Fatty Acids, Omega‐3"] 2360

#29 [mh ^"fish oils"] 1190

#30 [mh ^"cod liver oil"] 42

#31 [mh ^"docosahexaenoic acids"] 1415

#32 [mh ^"eicosapentaenoic acid"] 1227

#33 [mh ^"dietary fats, unsaturated"] 614

#34 [mh ^"alpha‐Linolenic Acid"] 256

#35 [mh ^"Dietary Supplements"] 13902

#36 (omega3* or omega 3*):ti,ab,kw 7585

#37 fatty acid*:ti,ab,kw 21333

#38 ((n3 or n‐3 or w3 or w‐3) NEAR/2 polyunsaturat*):ti,ab,kw 873

#39 (PUFA* or n‐3PUFA* or n3PUFA*):ti,ab,kw 2654

#40 ((n3 or n‐3 or w3 or w‐3) NEAR/2 oil*):ti,ab,kw 117

#41 (fish* NEAR/3 oil*):ti,ab,kw 3626

#42 (cod NEAR/2 oil*):ti,ab,kw 84

#43 (icosapentanoic* or icosapentaenoic* or timnodonic* or aan7qov9ea):ti,ab,kw 596

#44 (diet* NEAR/3 (fat* or oil*)):ti,ab,kw 11300

#45 (unsaturate* NEAR/2 (fat* or oil*)):ti,ab,kw 2013

#46 (alphalinolen* or alpha‐linolen*):ti,ab,kw 571

#47 (linolenate* or linolenic*):ti,ab,kw 1237

#48 Epanova:ti,ab,kw 34

#49 ((diet* or food*) NEAR/2 supplement*):ti,ab,kw 25512

#50 (Marine NEAR/2 (oil* or lipid*)):ti,ab,kw 123

#51 (Coromega or Efamed or MaxEPA or Lovaza or Omtryg or Vascepa or icosapent ethyl or DHA or Docosahex* or Eicosapent* or ethyl‐eicosapent* or ethyleicosapent* or EPA or bilantin omega or conchol 36 or eicosa e or eicosapen or sakana):ti,ab,kw 6031

#52 {OR #28‐#51} 53720

#53 #17 AND #27 AND #52 1737

CINAHL Plus with Full Text

Interface ‐ EBSCOhost Research Databases 
Search Screen ‐ Advanced Search

#	Query
S54	S11 AND S23 AND S53
S53	S24 OR S25 OR S26 OR S27 OR S28 OR S29 OR S30 OR S31 OR S32 OR S33 OR S34 OR S35 OR S36 OR S37 OR S38 OR S39 OR S40 OR S41 OR S42 OR S43 OR S44 OR S45 OR S46 OR S47 OR S48 OR S49 OR S50 OR S51 OR S52
S52	(MH “Dietary Supplements”)
S51	TI (Eicosapent* or ethyl‐eicosapent* or ethyleicosapent*) OR AB (Eicosapent* or ethyl‐eicosapent* or ethyleicosapent*)
S50	( TI (DHA) or AB (DHA) ) OR ( TI (EPA) OR AB (EPA) ) OR ( TI (Docosahex*) OR AB (Docosahex*) )
S49	TI (Vascepa or icosapent ethyl) OR AB (Vascepa or icosapent ethyl)
S48	( TI (Coromega) OR AB (Coromega) ) OR ( TI (Efamed) OR AB (Efamed) ) OR ( TI (MaxEPA) OR AB (MaxEPA) ) OR ( TI (Lovaza) OR AB (Lovaza) ) OR ( TI (Omtryg) OR AB (Omtryg) )
S47	TI (Marine N2 (oil* or lipid*)) OR AB (Marine N2 (oil* or lipid*))
S46	TI ((diet* or food*) N1 supplement*) OR AB ((diet* or food*) N1 supplement*)
S45	TI (Epanova) or AB (Epanova)
S44	TI (linolenate* or linolenic*) OR AB (linolenate* or linolenic*)
S43	TI (alphalinolen* or alpha‐linolen*) OR AB (alphalinolen* or alpha‐linolen*)
S42	TI (diet* N1 (fat* or oil*)) OR AB (diet* N1 (fat* or oil*))
S41	( TI (icosapentanoic*) OR AB (icosapentanoic*) ) OR ( TI (icosapentaenoic*) OR AB (icosapentaenoic*) ) OR ( TI (timnodonic*) OR AB (timnodonic*) ) OR ( TI (1553‐41‐9 or aan7qov9ea) OR AB (1553‐41‐9 or aan7qov9ea) )
S40	TI (cod N1 oil*) OR AB (cod N1 oil*)
S39	TI (fish* N1 oil*) OR AB (fish* N1 oil*)
S38	TI ((n3 or n‐3 or w3 or w‐3) N4 oil*) OR AB ((n3 or n‐3 or w3 or w‐3) N4 oil*)
S37	TI (PUFA* or n‐3PUFA* or n3PUFA*) OR AB (PUFA* or n‐3PUFA* or n3PUFA*)
S36	TI ((n3 or n‐3 or w3 or w‐3) N4 polyunsaturat*) OR AB ((n3 or n‐3 or w3 or w‐3) N4 polyunsaturat*)
S35	TI (“fatty acid*”) OR AB (“fatty acid*”)
S34	TI (omega3* or omega 3*) OR AB (omega3* or omega 3*)
S33	(MH “linolenic acids”)
S32	(MH “fatty acids, unsaturated”)
S31	(MH “dietary fats”)
S30	(MH “fatty acids”)
S29	(MH “alpha‐Linolenic Acid”)
S28	(MH “fats, unsaturated”)
S27	(MH “eicosapentaenoic acid”)
S26	(MH “docosahexaenoic acids”)
S25	(MH “fish oils”)
S24	(MH “Fatty Acids, Omega‐3”)
S23	S12 OR S13 OR S14 OR S15 OR S16 OR S17 OR S18 OR S19 OR S20 OR S21 OR S22
S22	TI ((grade* or form*) N1 (one or first or two or second or three or third or four* or five or fifth or six* or seven* or eight* or nine* or ninth or ten* or eleven* or twelfth or twelve or "1" or "2" or "3" or "4" or"5" or "6" or "7" or "8" or "9" or "10" or "11" or "12")) OR AB ((grade* or form*) N1 (one or first or two or second or three or third or four* or five or fifth or six* or seven* or eight* or nine* or ninth or ten* or eleven* or twelfth or twelve or "1" or "2" or "3" or "4" or"5" or "6" or "7" or "8" or "9" or "10" or "11" or "12"))
S21	TI (youngster* or preschool* or pre‐school* or kindergarten* or freshm#n* or junior* or soph#more* or senior* or highschool* or high school* or colleg* or universit* or student* or undergrad* or grade school* or elementary school* or middle school* or public school* or secondary school* or intermediate school*) OR AB (youngster* or preschool* or pre‐school* or kindergarten* or freshm#n* or junior* or soph#more* or senior* or highschool* or high school* or colleg* or universit* or student* or undergrad* or grade school* or elementary school* or middle school* or public school* or secondary school* or intermediate school*)
S20	TI (young* N1 (adult* or person* or individual* or people* or population* or man or men or wom#n)) OR AB (young* N1 (adult* or person* or individual* or people* or population* or man or men or wom#n))
S19	TI (minors or boy or boys or boyhood or girl or girls or girlhood or kid or kids or kiddie or child* or schoolchild* or adolescen* or preadolescen* or juvenil* or youth* or teen* or preteen* or underage* or under age* or pubescen* or prepubescen* or puberty or pe#diatric* or p#ediatric* or school*) OR AB (minors or boy or boys or boyhood or girl or girls or girlhood or kid or kids or kiddie or child* or schoolchild* or adolescen* or preadolescen* or juvenil* or youth* or teen* or preteen* or underage* or under age* or pubescen* or prepubescen* or puberty or pe#diatric* or p#ediatric* or school*)
S18	(MH "Adolescent Nutrition")
S17	(MH “Students, Elementary”) OR (MH “Schools, Middle”) OR (MH “Students, Middle School”) OR (MH “Schools, Secondary”) OR (MH “Students, High School”) OR (MH “High School Graduates”) OR ( (MH “Colleges and Universities”) ) OR (MH “Students, College”) OR (MH “Students, Undergraduate”)
S16	(MH “Minors (Legal)”) OR (MH “Puberty”)
S15	(MH “child”) OR (MH “child, preschool”)
S14	(MH "Adolescence")
S13	(MH “young adult”)
S12	(MH "Pediatrics")
S11	S1 OR S2 OR S3 OR S4 OR S5 OR S6 OR S7 OR S8 OR S9 OR S10
S10	TI (mental N2 (health or disorder* or illness* or stress or turmoil or despair)) OR AB (mental N2 (health or disorder* or illness* or stress or turmoil or despair))
S9	( TI (internaliz* or involutional) OR AB (internaliz* or involutional) ) OR ( TI (“affective disorder*” or “affective symptom*”) OR AB (“affective disorder*” or “affective symptom*”) )
S8	( TI (dysthymi*) or AB (dysthymi*) ) OR ( TI (melanchol*) or AB (melanchol*) ) OR ( TI (MDD) or AB (MDD) ) OR ( TI (mood*) or AB (mood*) )
S7	TI (depress*) or AB (depress*)
S6	(MH "Affective Symptoms")
S5	(MH "Premenstrual Dysphoric Disorder")
S4	(MH “Mental Disorders”)
S3	(MH "Affective Disorders, Psychotic")
S2	(MH "Dysthymic Disorder") OR (MH "Affective Disorders")
S1	(MH “Depression”)

Child Development & Adolescent Studies

Interface ‐ EBSCOhost Research Databases 
Search Screen ‐ Advanced Search

#	Query
S11	S3 AND S10
S10	S4 OR S5 OR S6 OR S7 OR S8 OR S9
S9	TI ( ((diet* or food*) N2 supplement*) ) OR AB ( ((diet* or food*) N2 supplement*) )
S8	TI ( (diet* or unsaturate*) N3 (fat* or oil*) ) OR AB ( (diet* or unsaturate*) N3 (fat* or oil*) )
S7	TI ( ((fish* OR cod) N3 oil*) ) OR AB ( ((fish* OR cod) N3 oil*) )
S6	TI ( PUFA* or n‐3PUFA* or n3PUFA* ) OR AB ( PUFA* or n‐3PUFA* or n3PUFA*). )
S5	TI ( (Coromega or Efamed or MaxEPA or Lovaza or Omtryg or Vascepa or icosapent ethyl or DHA or Docosahex* or Eicosapent* or ethyl‐eicosapent* or ethyleicosapent* or EPA or bilantin omega or conchol 36 or eicosa e or eicosapen or sakana) ) OR AB ( (Coromega or Efamed or MaxEPA or Lovaza or Omtryg or Vascepa or icosapent ethyl or DHA or Docosahex* or Eicosapent* or ethyl‐eicosapent* or ethyleicosapent* or EPA or bilantin omega or conchol 36 or eicosa e or eicosapen or sakana) )
S4	TI ( (omega3* or omega 3* or fatty acid* or icosapentanoic* or icosapentaenoic* or timnodonic* or 1553‐41‐9 or aan7qov9ea or alphalinolen* or alpha‐linolen* or linolenate* or linolenic* or Epanova) ) OR AB ( (omega3* or omega 3* or fatty acid* or icosapentanoic* or icosapentaenoic* or timnodonic* or 1553‐41‐9 or aan7qov9ea or alphalinolen* or alpha‐linolen* or linolenate* or linolenic* or Epanova) )
S3	S1 OR S2
S2	TI ( (mental N4 (health or disorder* or illness* or stress or turmoil or anguish or dispair)) ) OR AB ( (mental N4 (health or disorder* or illness* or stress or turmoil or anguish or dispair)) )
S1	TI ( (depress* OR dysthymi* OR melanchol* OR MDD OR mood* OR internaliz* or involutional OR affective disorder* or affective symptom OR dysphoric disorder*) ) OR AB ( (depress* OR dysthymi* OR melanchol* OR MDD OR mood* OR internaliz* or involutional OR affective disorder* or affective symptom OR dysphoric disorder*) )

Appendix 3. Grey literature search

Search Engine Searching

Date	Search engine	Search strategy(s) including how items were selected	# of items screened (uploaded to citation management software)
May 28, 2023	Google.ca	Omega‐3 fatty acids and depression in children Omega‐3 fatty acids and depression in adolescents Omega‐3 fatty acids and depression in children and adolescents Selection: Items were selected by scanning the first 100 results from each search	100 (0) 100 (0)
May 28, 2023	DuckDuckGo.com	Omega‐3 fatty acids and depression in children Omega‐3 fatty acids and depression in adolescents Omega‐3 fatty acids and depression in children and adolescents Selection: Items were selected by scanning the first 100 results from each search	100 (0) 100 (0)

Targeted Website Browsing/Searching

SEARCH TERMS

INTERVENTION: Omega‐3 fatty acids

CONDITION: depression OR mood OR depressive symptoms

POPULATION: Children OR adolescents OR pediatrics

Date	Organization Name	URL	# of items uploaded to citation management software
May 29, 2023	International Society for Nutritional Psychiatry Research	http://www.isnpr.org	0
May 29, 2023	Food & Mood Centre	foodandmoodcentre.com.au	0
May 29, 2023	Mental Health and Nutrition Research Group	https://www.nutritional-psychology.org/	0
May 29, 2023	NICM	http://nicm.edu.au/about_us/about_NICM	0
May 29, 2023	Health Canada	http://www.hc-sc.gc.ca/index-eng.php	0
May 29, 2023	Canadian Institute for health information	https://www.cihi.ca/en	0

Clinical Trials Databases Keywords: Omega 3 Fatty Acids AND depression AND children OR adolescents
Date	Organization name	URL	Keywords	# of results (# uploaded to citation management software)
May 28, 2023	WHO International Clinical Trials Registry Platform	https://trialsearch.who.int/Default.aspx	omega 3 fatty acids AND depress* (Selected option to search for clinical trial in children)	11 (0)
May 28, 2023	ClinicalTrials.gov	https://clinicaltrials.gov/	Condition/disease: depression Other terms: (none) Intervention/treatment: Omega 3 fatty acids Location: (no filter) Eligibility Criteria: Sex: All Age: Child (birth ‐ 17)	0 (0)
May 28, 2023	ISRCTN registry	https://www.isrctn.com/search?q=	Study status (no filter) Condition Category Mental and Behavioural Disorders Recruitment country (no filter) Age range Child Recruitment status (no filter) Results Status (no filter) Protocol Status (no filter)	45 (0)
May 28, 2023	Clinical Trials Listing Service	https://www.centerwatch.com/clinical-trials/listings	omega 3 fatty acids AND depress* (Selected option to search for clinical trial in children and adolescents 0‐18 years old)	0 (0)

Appendix 4. Version 2 of the Cochrane risk‐of‐bias tool for randomized trials (RoB 2)

Risk of bias arising from the randomization process

Low risk

• A random component was used in the sequence generation process (e.g. computer‐generated random numbers, reference to a random number table, coin‐tossing, shuffling cards or envelopes, throwing dice, or drawing lots). An allocation sequence that is generated using minimization will generally be considered random

• All the following

  • The allocation sequence was adequately concealed: the trial used any form of remote or centrally administered method to allocate interventions to participants (e.g. independent central pharmacy, telephone or Internet‐based randomization service providers) and envelopes or drug containers were used appropriately

  • Any baseline differences observed between intervention groups appear to be compatible with chance or there is no information about baseline imbalances

  • The allocation sequence was random or there is no information about whether the allocation sequence was random

High risk

• No random element was used in generating the allocation sequence or the sequence is predictable (e.g. methods based on dates, patient record numbers, allocation decisions made by clinicians or participants, allocation based on the availability of the intervention, any other systematic or haphazard method)

• Any of the following

  • The allocation sequence was not adequately concealed (there is reason to suspect that the enrolling investigator or the participant had knowledge of the forthcoming allocation)

  • There is no information about concealment of the allocation sequence and baseline differences between intervention groups suggest a problem with the randomization process (e.g. substantial differences between intervention group sizes, compared with the intended allocation ratio, a substantial excess in statistically significant differences in baseline characteristics between intervention groups, beyond that expected by chance; or imbalance in one or more key prognostic factors, or baseline measures of outcome variables, that is very unlikely to be due to chance and for which the between‐group difference is big enough to result in bias in the intervention effect estimate)

Some concerns

• Insufficient information about the sequence generation process to permit judgment

• The only information about randomization methods is a statement that the study is randomized

• There is no useful baseline information available

Risk of bias due to deviations from the intended interventions (effect of assignment to intervention)

Low risk

• Any of the following

  • Participants, carers, and people delivering the interventions were unaware of intervention groups during the trial (through the use of a placebo or sham intervention)

  • Participants, carers, or people delivering the interventions were aware of intervention groups during the trial and no deviations from intended intervention arose because of the trial context

• An appropriate analysis was used to estimate the effect of assignment to intervention

High risk

• All the following

  • Participants, carers, or people delivering the interventions were aware of intervention groups during the trial

  • There is evidence, or strong reason to believe, that the trial context led to failure to implement the protocol interventions or to implementation of interventions not allowed by the protocol

  • These deviations were likely to have affected the outcome

  • These deviations were unbalanced between the intervention groups

  • An appropriate analysis was not used to estimate the effect of assignment to intervention

  • The potential impact (on the estimated effect of the intervention) of the failure to analyze participants in the group to which they were randomized was substantial

Some concerns

• Insufficient information about the deviations from the intended interventions to permit judgment

Risk of bias due to deviations from the intended interventions (effect of adhering to intervention)

Low risk

• An appropriate analysis was used to estimate the effect of assignment to intervention

• All the following

  • Participants, carers, and people delivering the interventions were unaware of intervention groups during the trial or participants, carers, or people delivering the interventions were aware of intervention groups and the important non‐protocol interventions were balanced across intervention groups

  • Failures in implementing the intervention could not have affected the outcome

  • Study participants adhered to the assigned intervention regimen

High risk

• Participants experienced adverse effects or toxicities that they knew to be specific to one of the interventions

• Participants experienced adverse effects or toxicities that carers or people delivering the interventions knew to be specific to one of the interventions​​​​​​

• Any of the following

  • Participants, carers, and people delivering the interventions were unaware of intervention groups during the trial and failures in implementing the intervention could have affected the outcome or study participants did not adhere to the assigned intervention regimen

  • Participants, carers, or people delivering the interventions were aware of intervention groups and the important non‐protocol interventions were balanced across intervention groups. Failures in implementing the intervention could have affected the outcome or study participants did not adhere to the assigned intervention regimen

  • Participants, carers, or people delivering the interventions were aware of intervention groups and the important non‐protocol interventions were not balanced across intervention groups

• An appropriate analysis was not used to estimate the effect of adhering to intervention

Some concerns

• Insufficient information about deviations from the intended interventions to permit judgment

Risk of bias due to missing outcome data

Low risk

• Analysis methods that correct for bias or sensitivity analyses showing that results are little changed under a range of plausible assumptions about the relationship between missingness in the outcome and its true value

• Outcome data were available for all, or nearly all, randomized participants

• All missing outcome data occurred for documented reasons that are unrelated to the outcome

• There is evidence that the result was not biased by missing outcome data

• Missingness in the outcome could not depend on its true value

• The analysis accounted for participant characteristics that are likely to explain the relationship between missingness in the outcome and its true value

High risk

• Outcome data were not available for all, or nearly all, randomized participants

• There is no evidence that the result was not biased by missing outcome data

• Missingness in the outcome could depend on its true value

• Missingness in the outcome likely depended on its true value

• If loss to follow‐up, or withdrawal from the study, could be related to participants' health status

• Differences between intervention groups in the proportions of missing outcome data. If there is a difference between the effects of the experimental and comparator interventions on the outcome, and the missingness in the outcome is influenced by its true value, then the proportions of missing outcome data are likely to differ between intervention groups. Such a difference suggests a risk of bias due to missing outcome data because the trial result will be sensitive to missingness in the outcome being related to its true value. For time‐to‐event‐data, the analogue is that rates of censoring (loss to follow‐up) differ between the intervention groups

• Reported reasons for missing outcome data provide evidence that missingness in the outcome depends on its true value

• Reported reasons for missing outcome data differ between the intervention groups

• The circumstances of the trial make it likely that missingness in the outcome depends on its true value (i.e. continuing symptoms make drop out more likely)

• In time‐to‐event analyses, participants' follow‐up is censored when they stop or change their assigned intervention, for example, because of toxicity

Some concerns

• Insufficient information about missing outcome data to permit judgment

Risk of bias in the measurement of the outcome

Low risk

• Both of the following

  • The method of measuring the outcome was appropriate and the measurement or ascertainment of the outcome did not differ between intervention groups

  • The outcome assessors were unaware of the intervention received by study participants or the assessment of the outcome could not have been influenced by knowledge of the intervention received

High risk

• Outcome assessors were blinded to intervention status. For participant‐reported outcomes, the outcome assessor is the study participant.

• Any of the following

  • The method of measuring the outcome was inappropriate

  • The measurement or ascertainment of the outcome could have differed between intervention groups

  • It is likely that the assessment of the outcome was influenced by knowledge of the intervention received

Some concerns

• Insufficient information about the measurement of the outcome process to permit judgment

Risk of bias in the selection of the reported result

Low risk

• The researchers' prespecified intentions are available in sufficient detail

• Changes to analysis plans were made before unblinded outcome data were available, or were clearly unrelated to the results

• There is clear evidence that all eligible reported results for the outcome measurement correspond to all intended analyses. There is only one possible way in which the outcome measurement can be analyzed (hence there is no opportunity to select from multiple analyses)

• Analyses are inconsistent across different reports on the same trial, but the trialists have provided the reason for the inconsistency and it is not related to the nature of the results

• All of the following

  • The data were analyzed in accordance with a prespecified plan that was finalized before unblinded outcome data were available for analysis

  • The result being assessed is unlikely to have been selected, on the basis of the results, from multiple eligible outcome measurements (e.g. scales, definitions, time points) within the outcome domain

  • Reported outcome data are unlikely to have been selected, on the basis of the results, from multiple eligible analyses of the data

High risk

• There is clear evidence that measurement was analyzed in multiple eligible ways, but data for only one or a subset of analyses is fully reported (without justification), and the fully reported result is likely to have been selected on the basis of the results

• Selection on the basis of the results arises from a desire for findings to be newsworthy, sufficiently noteworthy to merit publication or to confirm a prior hypothesis

• One of the following

  • The result being assessed is likely to have been selected, on the basis of the results, from multiple eligible outcome measurements (e.g. scales, definitions, time points) within the outcome domain

  • The result being assessed is likely to have been selected, on the basis of the results, from multiple eligible analyses of the data

Some concerns

• Any analysis intentions are not available, or the analysis intentions are not reported in sufficient detail to enable an assessment, and there is more than one way in which the outcome measurement could have been analysed

• Insufficient information about the selection of the reported result process to permit judgment

Data and analyses

Comparison 1. Primary outcomes.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1.1 Self‐reported depression symptoms (continuous measure)	5	185	Std. Mean Difference (IV, Random, 95% CI)	‐0.34 [‐0.85, 0.17]
1.2 Depression remission (dichotomous measure)	4	127	Odds Ratio (M‐H, Random, 95% CI)	1.11 [0.45, 2.75]

Comparison 2. Secondary outcomes.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
2.1 Attrition (dropouts)	5	228	Odds Ratio (IV, Random, 95% CI)	0.94 [0.46, 1.90]

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Fristad 2019.

Study characteristics
Methods	Randomized controlled pilot study
Participants	Participants: 72 participants aged 7–14 years (mean age 11.6 [SD 2.1]), 43% female. Participants were recruited from community advertisements and clinician referrals. Comorbidities: common comorbid disorders included anxiety, ADHD, and disruptive behavior disorders Adjunctive treatment: none Inclusion criteria: aged 7–14 years; diagnosis of MDD, dysthymic disorder, or depressive disorder; clinically significant symptom severity on the CDRS‐R scores (≥ 40); youth and ≥ 1 caregiver completed the screening assessment and were willing and able to participate in follow‐up procedures. Exclusion criteria: inability to swallow capsules; major medical disorder; autistic disorder; psychosis, antipsychotic medication; active suicidal concern; mental health intervention (psychotherapy or pharmacotherapy other than medication for ADHD or a sleep aid) or omega 3 in the month preceding randomization; intellectual disability (IQ < 70 and impaired adaptive functioning); lack of access to a telephone.
Interventions	Intervention Experimental arm 1: omega‐3 supplementation: omega‐3 PUFAs 1000 mg (2 × 500 mg capsules, each containing EPA 350 mg and DHA 50 mg; 100 other omega‐3 PUFA) by mouth, twice daily and psychoeducational psychotherapy sessions occur twice a week for up to 24 sessions. Experimental arm 2: omega‐3 (same as above) and no psychoeducational psychotherapy Experimental arm 3: placebo supplement and psychoeducational psychotherapy Comparator:placebo comparator: placebo supplement contained no active fatty acids. 2 placebo capsules twice daily (matched omega‐3 PUFA for odor and appearance) Duration: 12 weeks
Outcomes	Primary outcome Change in depressive symptom severity: continuous scale using CDRS‐R at week prior to randomization and then weeks 2, 4, 6, 9, and 12 after randomization Secondary outcomes Adverse events: monitored but there were no differences between intervention and placebo groups Failure to complete: 2 participants receiving omega‐3 PUFA were lost to follow‐up Compliance measured by pill counts and no differences between intervention and placebo groups
Notes	Study funding: National Institute of Mental Health. There were no restrictions placed by the funding sources on the reporting of findings. Supplements provided by OmegaBrite Corporation Conflicts of interest: none

Gabbay 2018.

Study characteristics
Methods	Randomized controlled parallel‐arm study
Participants	Participants: 51 participants aged 12–19 years (mean age 16.1 [SD 2.1] years). 57% female Comorbidities: ADHD Adjunctive treatment: none, psychotropic medication–free Inclusion criteria: adolescents aged 12–19 years with a diagnosis of current MDD (according to DSM‐IV‐TR) of ≥ 6 weeks' duration, and CDRS‐R minimum raw score of 40, IQ > 80, and ability to swallow pills Exclusion criteria: significant medical or neurologic disorders; allergy to seafood/fish; lifetime diagnoses of bipolar disorder, autism, pervasive developmental disorder, or Tourette's disorder; and current diagnoses of eating disorder, panic disorder, obsessive‐compulsive disorder, post‐traumatic stress disorder, conduct disorder, and substance‐related disorder, other than nicotine dependence
Interventions	Intervention: fixed‐flexible dose titration schedule end mean dose 3.4 g/day; EPA:DHA ratio 2:1. The fixed flexible dose titration schedule was as follows: initial dose 1.2 g/day was increased 0.6 g/day every 2 weeks, up to a maximum of 3.6 g/day. Comparator: placebo capsules identical in appearance and scent/taste to intervention capsules. Placebo capsules contained a 50:50 ratio of corn and soybean oils, omega‐6 (50%), monounsaturated (25%) fatty acids. Duration: 10 weeks
Outcomes	Primary outcomes Depression severity measured as a continuous variable as assessed by clinician‐rated CDRS‐R and self‐reported BDI‐II. Treatment response, assessed at end of treatment, primarily defined as ≥ 50% improvement from baseline on the CDRS‐R total score after subtracting the 17‐item base score, or reaching a CDRS‐R score ≤ 28. Categorical variable reported as very much improved or much improved. Secondary outcomes Adverse events: routinely monitored (none reported) Failure to complete: 3 participants in intervention group and 6 participants in placebo group did not complete the study. Compliance measured by asking parents/guardians to supervise each instance of dosing. Adolescents were also asked to return medication bottles, and capsule counts were performed at each weekly visit.
Notes	Study funding: National Institute of Health. There were no restrictions placed by the funding sources on the reporting of findings. Ocean Nutrition Canada provided the supplements. Conflicts of interest: none reported

McNamara 2020.

Study characteristics
Methods	Randomized, double‐blind, parallel‐group, placebo‐controlled fixed‐dose trial
Participants	Participants: 56 participants aged 9–21 years (mean age 14.1 [SD 3.0] years). 80% female Comorbidities: ADHD, anxiety Adjunctive treatment: no Inclusion criteria: adolescents aged 9–21 years with a current DSM‐IV diagnosis of MDD, or depressive disorder not otherwise specified, CDRS‐R raw score ≥ 40, and ≥ 1 biologic parent with bipolar disorder, type I, as determined by the Structured Clinical Interview for DSM Disorders Exclusion criteria: mood symptoms resulting from acute medical illness or acute intoxication or withdrawal from drugs or alcohol; psychotic symptoms; lifetime history of a manic or hypomanic episode; lifetime diagnosis of bipolar disorder or cyclothymia; history of a major medical, neurologic illness, laboratory abnormalities, or a significant episode (> 10 minutes) of loss of consciousness; history of alcohol or substance dependence (nicotine dependence was permitted); intellectual disability (IQ < 70); positive serum pregnancy test; intolerance, hypersensitivity, or non‐response to omega‐3 fatty acids; shellfish or seafood allergy; history of a clotting disorder in themselves or a first‐degree relative; concomitant use of medications with anticoagulant effects; and use of antipsychotics, other mood stabilizers, stimulants (if opting to discontinue), or atomoxetine within 72 hours (aripiprazole within 2 weeks was exclusionary because of its long half‐life) or antidepressants within 5 days (fluoxetine within 1 month was exclusionary because of its long half‐life). Concomitant use of other psychotropic medications or medications with central nervous system effects within 5 half‐lives from baseline or prior treatment with a medication with central nervous system effects that requires > 5 days of a screening period; requiring psychiatric admission, as determined by the study physician; and initiation of psychotherapy within 2 months of the screening visit, or plans to initiate psychotherapy during study participation. People diagnosed with ADHD and taking a stable dose of stimulants for the previous month were permitted to continue.
Interventions	Intervention: 3 fish oil capsules daily with EPA 450 mg, DPA 40 mg, and DHA 260 mg for a total daily dose of EPA + DHA 2130 mg Comparator: placebo identical in size, shape, and color. 3 placebo capsules. Each placebo capsule contained linoleic acid 92 mg, stearic acid 23 mg, palmitic acid 114 mg, and oleic acid 746 mg Duration: 12 weeks
Outcomes	Primary outcome Change in depressive symptom severity on a continuous scale using CDRS‐R Secondary outcomes Adverse events: monitored but no differences between groups observed Failure to complete: 6 participants receiving omega‐3 PUFA and 8 in the placebo group did not complete the trial Compliance was not measured by pill counts; rather, fasting serum laboratory test at baseline and endpoint of erythrocyte fatty acid composition (mg fatty acid/100 mg fatty acids) was determined by gas chromatography to measure EPA + DHA, as well as omega‐3 PUFA (EPA + DPA + DHA), arachidonic acid (20:4n‐6), the arachidonic acid/EPA + DHA ratio.
Notes	Study funding: National Institute of Health. There were no restrictions placed by the funding sources on the reporting of findings. Supplements supplied by Inflammation Research Foundation Conflicts of interest: 3 authors reported conflicts of interest.

Nemets 2006.

Study characteristics
Methods	A controlled, double‐blinded pilot study
Participants	Participants: 28 participants aged 6–12 years (mean age 10.1 years). 25% female. Participants recruited from a hospital depression clinic Comorbidities: anxiety, ADHD, dysthymia, or tic syndrome Adjunctive treatment: no Inclusion criteria: aged 6–12 referred to the depression clinic at the Schneider Children's Medical Center of Israel or the Child Psychiatry Clinic of the Beer‐Sheva Mental Health Center. Children had MDD diagnosed using the Hebrew translation of the childhood version of the Schedule for Affective Disorders and Schizophrenia Exclusion criteria: unstable physical illness or psychiatric disorders other than anxiety, ADHD, dysthymia, or tic syndrome
Interventions	Intervention: 2 × 500 mg capsule (EPA 190 mg + DHA 90 mg) or 1 × 1000 mg capsule (EPA 400 mg + DHA 200 mg) daily (mean dose 1 g/day) Comparator: placebo for the 500 mg capsule was olive oil. Placebo for 1000 mg capsules was safflower oil. Duration: 16 weeks
Outcomes	Primary outcome Change in depressive symptom severity on a continuous scale using CDRS‐R Secondary outcomes Adverse events: monitored but none reported Failure to complete: 3 participants receiving omega‐3 PUFA and 5 in the placebo group did not complete the trial Compliance: not reported
Notes	Study funding: National Alliance for Research in Schizophrenia and Depression Junior Investigator Award. There were no restrictions placed by the funding sources on the reporting of findings. Ocean Nutrition Canada provided the omega‐3 PUFA supplements; Sears Laboratories provided the placebo Conflicts of Interest: none

Trebaticka 2020.

Study characteristics
Methods	Single‐center, randomized, double‐blind controlled pilot study
Participants	Participants: 38 participants aged 11–17 years (mean age 15.6 [SD 1.6]). 80% female. Participants were recruited from a hospital depression clinic Comorbidities: none Adjunctive treatment: typical antidepressant therapy Inclusion criteria: diagnosis of depressive disorder or mixed anxiety and depressive disorder (according to International Classification of Diseases, 10th Edition), age 7–18 years, with no indication of chronic somatic disease and normal eating habits. Exclusion criteria: chronic somatic diseases, dietary restrictions, psychotic disorders, eating disorders, addiction to psychoactive compounds, personality disorders, organic mental disorders, and pervasive developmental disorders
Interventions	Intervention: 20 mL of omega‐3 fish oil emulsion (providing total omega‐3 fatty acid 2400 mg; EPA 1000 mg and DHA 750 mg, EPA:DHA ratio 1.33:1) Comparator: identical‐looking comparator was an omega‐6 fatty acid‐rich sunflower oil emulsion containing omega‐6 linoleic acid 2467 mg Duration: 12 weeks
Outcomes	Primary outcome Change in depressive symptom severity on a continuous scale using CDI score, MADD score Secondary outcomes Serious adverse effects: monitored but none reported Failure to complete: 3 participants receiving omega‐3 PUFA and 5 in the placebo group did not complete the trial Compliance: product returns were within expected volume, exceeding 95% compliance.
Notes	Study funding: partly financially supported by the VEGA grant 01/0703/13 of Ministry of Education of SR, APVV Grant 15‐0063 and Mind and Health, civil association. There were no restrictions placed by the funding sources on the reporting of findings. Supplements (emulsion oils) were provided by Cultech Ltd Conflicts of interest: 1 author declared a conflict of interest

ADHD: attention deficit hyperactivity disorder; BDI‐II: Beck Depression Inventory II; CDI: Children's Depression Inventory; CDRS‐R: Children's Depression Rating Scale–Revised; DHA: docosahexaenoic acid; DPA: docosapentaenoic acid; DSM‐IV: Diagnostic and Statistical Manual of Mental Disorders Fourth Edition; DSM‐IV‐TR: Diagnostic and Statistical Manual of Mental Disorders Fourth Edition, Text Revision; EPA: eicosapentaenoic acid; IQ: intelligence quotient; MADD: Mixed Anxiety and Depressive Disorder; MDD: major depressive disorder; PUFA: polyunsaturated fatty acid; SD: standard deviation.

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Cadenhead 2017	Adult population
Carney 2019	Adult population
Chan 2015	Depression not measured
Chhetry 2016	Adult population
Clayton 2007	Not a randomized controlled trial
Cornu 2018	Depression not measured
Crowe 2007	Not a randomized controlled trial
CTRI/2020/10/028383	Adult population
Domenech 2015	Not a randomized controlled trial
Gholami 2021	Depression not measured
Ginty 2015	Adult population
Hallahan 2007	Adult population
Hysing 2018	Depression not measured
Kennedy 2009	Depression not measured
Kiecolt‐Glaser 2011	Adult population
Kooshki 2014	Adult population
Lesperance 2005	Adult population
Marangell 2003	Adult population
McGorry 2017	Depression not measured
Meyer 2013	Adult population
NCT00511810	Adult population
NCT01104194	Adult population
NCT01341925	Adult population
NCT03732378	Adult population
Nelson 2018	Adult population
Park 2015	Adult population
Raine 2015	Depression not measured
Rapaport 2013	Adult population
Rapaport 2016	Adult population
Rapaport 2020	Adult population
Robinson 2019	Adult population
Rodriguez‐Hernandez 2020	Depression not measured
Salem 2019	Not a randomized controlled trial
Schuchardt 2008	Not a randomized controlled trial
Silvers 2005	Adult population
Smith 2018	Adult population
Su 2003	Adult population
Tyszkiewicz 2009	Not a randomized controlled trial
van der Wurff 2016	Depression not measured
van Meter 2017	Depression not measured
Vargas 2015	Not a randomized controlled trial
Wozniak 2015	Depression not measured
Yang 2022	Adult population
Young 2019	Depression not measured

Characteristics of ongoing studies [ordered by study ID]

Cao 2021.

Study name	Study on the effect of omega‐3 polyunsaturated fatty acids in adolescents with depression
Methods	Design: double‐blind parallel study design Sample size: experimental group: 130; control group: 130
Participants	Inclusion criteria: males and females aged 13–25 years; based on screening for Structured Clinical Interview for DSM‐5, DSM‐5 was diagnosed as major depressive disorder or bipolar disorder; understand the content of the study, be willing to participate in the study, and sign the written informed consent Exclusion criteria: other mental disorders that meet DSM‐5 diagnostic criteria include organic mental disorders, alcohol dependence, drug dependence/abuse, and schizophrenia; have active suicidal ideation or plan to commit suicide, or have committed suicide in the last 6 months, or combinations of these; have severe psychotic symptoms; are taking supplements rich in omega‐3 PUFAs or taking at least 1 g of omega‐3 PUFAs daily; are taking > 400 IU of vitamin E supplements; clinically significant, unstable physical diseases; smoking > 10 cigarettes a day; positive pregnancy test or breastfeeding or planned to become pregnant during the course of the study.
Interventions	Intervention: omega‐3 PUFAs Comparator: corn oil intervention
Outcomes	Primary outcome: Hamilton Depression Rating Scale – 24‐items Secondary outcome: erythrocyte membrane fat ratio; brain structure and function information
Starting date	1 March 2021
Contact information	coolliping@163.com
Notes	The response of antidepressants and mood stabilizers for adolescents is different from that of adults. Many antidepressants and mood stabilizers for adults are not suitable for the treatment of adolescent mood disorders because their effectiveness and safety are not good. While ω‐3‐PUFAs have been shown to be effective in improving mood symptoms in adults with mood disorders, it remains unclear whether they have the same effect in adolescents. Therefore, it is necessary to conduct clinical studies to confirm the efficacy of ω‐3‐PUFAs in the treatment of adolescents with mood disorders.

Haberling 2019.

Study name	Omega‐3 fatty acids as a treatment for pediatric depression
Methods	Design: phase III, 36‐week, multicenter, double‐blind, placebo‐controlled randomized superiority study Sample size: 220 participants
Participants	Inclusion: male or female in‐ or outpatients of a participating center; children aged 8 to < 13 years or teenagers aged 13 to < 18 years; major depressive disorder with depressive symptoms of at least moderate severity; written informed consent of the parents/legal representatives and participant's assent Exclusion: contraindications to the drug; > 4 weeks of regular omega‐3 supplementation; pregnant or breastfeeding or intention to become pregnant; pre‐existing neurologic or medical conditions likely to be responsible for depressive symptoms; laboratory screening values considered clinically relevant; known or suspected non‐compliance; other psychiatric diagnoses (substance dependency, schizophrenia, bipolar affective disorder, eating disorder, mental retardation, pervasive developmental disorder); inability to follow the procedures of the study; participation in another study with omega‐3, previous enrolment in the current study, or dependent people of the investigators
Interventions	Intervention: daily dose of EPA 1000 mg/DHA 500 mg in children aged > 13 years. In children aged < 13 years, half the dose is administered resulting in EPA 500 mg/DHA 250 mg per day Comparator: placebo capsules containing mostly medium chain triglycerides, with a small amount of fish oil added to mimic the fishy taste and prevent guessing of the treatment arm
Outcomes	Primary outcome: change of the (continuous) Children's Depression Rating Scale‐revised (CDRS‐R) total score Secondary outcomes: response rate; remission rate; correlation between omega‐3 index and the overall score of the Scale of Impulsivity and Emotion Dysregulation (IES‐27); antidepressant medication; Children's Global Assessment Scale (CGAS); Kidscreen quality of life measure for children and adolescents; hospitalization; retention rate; inflammatory mediators as predictors of response; metabolites of EPA as predictors for response; depressive symptoms; Suicidal Ideation Questionnaire (SIQ); relationship between stress, omega‐3 fatty acids and saliva cortisol; recovery rate, cognition
Starting date	Recruitment completed June 2021, study underway.
Contact information
Notes	Correspondence with primary author May 2023 indicated that the study data have not been completed.

Li 2022.

Study name	Effects of ω‐3 polyunsaturated fatty acid supplementation in the treatment of adolescents with depressive disorder
Methods	Design: single‐center interventional, 12‐week open‐label randomized controlled trial Computer‐generated random sequence using block randomization with random block sizes. An independent statistician performed the randomization. Participants are enrolled and assigned sequentially to adjuvant interventions by the physician. The allocation sequence is not available to any member of the research team until the databases had been completed and locked. Sample size: 71 participants (34 in intervention group, 37 in comparator group)
Participants	Inclusion criteria: meet the diagnostic criteria of International Classification of Diseases 10th Edition (ICD‐10) for depressive disorder; aged 13–24 years; never received antidepressant medication, or have not received antidepressant medication within 2 weeks; person's family members or legal guardians fully understood the content of this study and signed informed consent Exclusion criteria: people diagnosed with bipolar disorder or other according to the ICD‐10; recent or past history of alcohol and drug dependence, other psychiatric disorders, eating disorders, or intellectual disability; serious or chronic physical disease, endocrine disease, or organic brain diseases; regular consumption of ω‐3 PUFA supplements or eating fish ≥ 4 times per week; conducting additional clinical trials within 4 weeks prior to the start of this study
Interventions	Intervention: ω‐3 PUFA rich fish oil + paroxetine: ω‐3 PUFA 2700 mg; EPA 1941 mg and DHA 759 mg; EPA:DHA ratio 2.56:1, paroxetine 20 mg/day Comparator: paroxetine 20 mg/day
Outcomes	Primary outcome: depressive symptoms measured using the Montgomery‐Osberg Depression Rating Scale at baseline, weeks 4, 8, and 12 Secondary outcomes: cognition, including memory conditions; niacin skin flushing response; plasma metabolome; membrane lipids
Starting date	29 November 2020
Contact information	230204199609291726@sjtu.edu.cn
Notes	Recruitment underway

Rice 2016.

Study name	Youth depression alleviation: the Fish Oil Youth Depression Study (YoDA‐F)
Methods	Design: prospective, randomized, double‐blind, placebo‐controlled, 12‐week PUFA intervention Sample size: 400 participants (200 per arm: allocated to treatment with either omega‐3 PUFAs plus cognitive behavioral case management (CBCM) or to treatment with placebo plus CBCM)
Participants	Inclusion criteria: age 15–25 at presentation; seeking help for psychological distress; moderate‐to‐severe depression, indicated by a score of ≥ 11 and ≤ 20 on the Quick Inventory of Depression Symptomatology, Adolescent Version (QIDS‐A17‐C) at first contact with the service, and again after approximately 1 week; diagnosis of major depressive disorder (MDD) at baseline (i.e. when the second QIDS‐A17‐C is completed) using a structured clinical interview; ability to give informed consent (for individuals aged < 18 years, written informed consent of ≥ 1 parent or a guardian is required) Exclusion criteria: lifetime diagnosis or history of treatment for psychotic disorder or bipolar disorder or substance dependence; history of treatment with an antidepressant (> 4 weeks during the last 12 months); acute suicidal behavior or aggressive behavior (indicated by a Comprehensive Assessment of At‐Risk Metal State (CAARMS) 70 score of 6 on the relevant items); depression secondary to a medical condition; intelligence quotient < 70; pregnancy (indicated by blood test) or lactation; laboratory values > 15% outside the normal range for bleeding parameters; current (or recent; within 8 weeks of being included in the trial, for > 1 week) use of omega‐3 supplements or psychotropic medication; known allergy of omega‐3 supplements
Interventions	Daily dose of 4 gelatine capsules (participants will be instructed to take 2 in the morning and 2 at night; before meals) throughout the 12‐week treatment period Intervention: daily dose of 4 capsules provides 0.650–0.750 g concentrated marine fish oil, participants who are randomized to receive fish oil will be taking a total of approximately 1.4 g per day of omega‐3 PUFAs Comparator: daily dose of 4 capsules provides 0.650–0.750 g of paraffin oil. To ensure blinding, placebo capsules will be carefully matched in appearance and flavor with the active treatment; they will also contain the same amount of vitamin E as the fish oil capsules, and approximately 1% fish oil to mimic taste.
Outcomes	Primary: change in depressive symptoms from baseline to 12 weeks, which will be assessed with the QIDS‐A17‐C Secondary: change in depressive symptoms at 26‐week follow‐up, remission rate (defined as participants not meeting diagnostic criteria for major depressive disorder), changes to biochemical assessments, and changes to symptomology and psychosocial functioning assessed across a range of domains
Starting date	May 2014
Contact information
Notes	Correspondence with primary author on 2 June 2021 indicated that the results are not yet published.

DHA: docosahexaenoic acid; DSM‐4: Diagnostic and Statistical Manual of Mental Disorders Fourth Edition; DSM‐5: Diagnostic and Statistical Manual of Mental Disorders Fifth Edition; EPA: eicosapentaenoic acid; PUFA: polyunsaturated fatty acid; SD: standard deviation.

Differences between protocol and review

We made the following changes from the published protocol (Campisi 2021b).

Secondary objectives

We added the following secondary objectives.

• To determine compliance and attrition for omega‐3 PUFAs as an intervention for depression in children and adolescents

• To determine the safety of omega‐3 PUFAs as an intervention for depression in children and adolescents

Participants

We changed the age criteria to 19 years or younger rather than between 6 and 19 years of age to be more inclusive and consistent with our search strategy. The search strategy did not exclude studies of children aged under 6 years. Additionally, we did not exclude any studies using 'under 6 years of age' as an eligibility criterion.

Study designs

None of the included studies were cross‐over or cluster‐RCTs. Future updates should follow the methods outlined below.

Cross‐over randomized controlled trials

In the case of the inclusion of a cross‐over RCT, we will treat the control and intervention periods as separate entities and extract data from the first period (before cross‐over) only.

Cluster‐randomized controlled trials

If we include a cluster‐RCT, we will calculate a cluster mean and use that summary measure with the cluster as the unit of analysis.

Subgroup analysis and meta‐regression

We originally planned to investigate whether various study characteristics could explain the clinical diversity/heterogeneity, but the number of studies included failed to meet the threshold (10 included studies) for subgroup analysis and meta‐regression. The following subgroup analyses and meta‐regression analyses were not pursued.

• Type omega‐3 PUFA

  • Studies where only DHA was used

  • Studies where only EPA was used

  • Studies where a combination was used

• Adjunctive therapy

  • Studies involving individuals receiving adjunctive therapies

  • Studies involving individuals not receiving adjunctive therapies

  • Studies involving a mix of individuals receiving and not receiving adjunctive therapies

• Age (in years)

• Duration of omega‐3 PUFA supplementation (in weeks)

• Dose levels of omega‐3 PUFAs (in milligrams)

• Duration of follow‐up (in weeks)

• Percentage of female participants

For future updates, we plan to also pursue subgroup analysis by types of comparators (including placebo, wait list controls, no treatment/supplementation, or standard care) to assess differences in effect size due to the use of different comparators.

Contributions of authors

SCC: conceptualized the systematic review

DJK, PZ, GB‐R, CZ, and SCC: contributed to the design and protocol of the systematic review

GB‐R: developed the search strategy and carried out the search

CZ and SCC: screened title, abstracts, and full‐text articles

CZ, AS, and SCC: performed the data extraction

SCC and AS: developed the analysis plan and conducted the analysis

CZ and SCC: drafted the initial manuscript

PZ, GB‐R, and DJK: provided critical feedback on the analysis and contributed to the writing process

All authors: read and approved the final manuscript.

Sources of support

Internal sources

• Hospital for Sick Children, Canada

  Susan C Campisi is supported in part by the SickKids RestraComp Post Doctoral Award and by the Canadian Institutes of Health Research (CIHR) grant number 409491.

External sources

• Cundill Centre for Child and Youth Depression at the Centre for Addiction and Mental Health, Canada

  Funding

Declarations of interest

SCC: none.

CZ: none.

GB‐R: none.

AS: none.

PS: none.

DJK: Canadian Academy of Child and Adolescent Psychiatry (Chair, Research and Scientific Program Committee); Canadian Pediatric Society (Chair, Mental Health Task Force); Canadian Institutes of Health Research (CIHR) (Grant/Contract); Hospital for Sick Children (Grant/Contract); University of Toronto (Grant/Contract); Psychiatrist, Hospital for Sick Children.

New