Efficacy of psychological interventions for young adults with mild-to-moderate depressive symptoms: A meta-analysis

(1)

Journal of Psychiatric Research 152 (2022) 366–374

Available online 23 June 2022

Efficacy of psychological interventions for young adults with mild-to-moderate depressive symptoms: A meta-analysis

J.C. Medina

^a^,^b^,^*

, C. Paz

^c

, H. García-Mieres

^d^,^e

, N. Ni no-Robles ˜

^b

, J.E. Herrera

^c

, G. Feixas

^b^,^f

, A. Montesano

^a

aDepartment of Psychology and Education Sciences, Universitat Oberta de Catalunya, Barcelona, Spain

bDepartment of Clinical Psychology and Psychobiology, Universitat de Barcelona, Barcelona, Spain

cSchool of Psychology, Universidad de Las Am´ericas, Quito, Ecuador

dEtiopathogenesis and Treatment of Severe Mental Disorders (MERITT), Teaching, Research & Innovation Unit, Institut de Recerca Sant Joan de D´eu, Parc Sanitari Sant Joan de D´eu, Sant Boi de Llobregat, Barcelona, Spain

eCentro Investigaci´on Biom´edica en Red Salud Mental (CIBERSAM), Madrid, Spain

fInstitute of Neurosciences, Universitat de Barcelona, Barcelona, Spain

A R T I C L E I N F O Keywords:

Psychotherapy Clinical psychology Affective symptoms Mood disorders Youth

A B S T R A C T

Background: Psychological interventions are commonly used to treat mild-to-moderate depression, but their efficacy in young adults has not been exhaustively addressed. This meta-analysis aims to establish it in comparison to no treatment, wait-list, usual treatment, passive interventions, and other bona-fide treatments.

Methods: The search was conducted in Scopus, MEDLINE, PsycINFO, ClinicalTrials.gov, the ISRCTN Registry, Cochrane CENTRAL, Clarivate BIOSIS Previews and the METAPSY database, retrieving studies from the start of records to April 2020. Eligibility criteria included samples of 16–30 years experiencing mild-to-moderate depressive symptoms and participating in randomized controlled trials (RCTs), non-RCTs, or pre-post studies measuring depressive symptomatology and featuring psychological treatments.

Results: Up to 45 studies met criteria, consisting of 3,947 participants, assessed using the Quality Assessment Tool for Quantitative Studies and their results meta-analyzed assuming random effects. Psychological interventions proved to be efficacious in RCTs compared to no treatment (g = −0.68; 95% CI = − 0.87, −0.48) and wait-list (g

= −1.04; 95% CI = −1.25, −0.82), while depressive symptoms also improved in pre-post studies (g = −0.99;

95% CI = −1.32, − 0.66). However, intervention efficacy was similar to usual care, passive, and bona-fide comparators. The heterogeneity found, a likely reporting bias and the low quality of most studies must be considered when interpreting these results.

Conclusions: Psychological treatments are efficacious to reduce depressive symptoms in young adults, but comparable to other interventions in the mild-to-moderate range. Moderators like depression severity or therapist involvement significantly influenced their efficacy, with results encouraging clinicians to adopt flexible and personalized approaches.

1. Introduction

Young or emerging adulthood is a critical period of development. It is a stage of transition from school to work, from being dependent on parents to attain individual responsibilities, and from occasional dating to establishing stable relationships (see Institute of Medicine and Na- tional Research Council of the National Academies, 2015). The age range that embraces this stage is not well-defined as it is influenced by the cultural and legal framework from each region. It is described for

some as 15–24 years, while others extend it to 30 years (The Society for Adolescent Health and Medicine, 2017), but what is clear is that young adults face unique challenges that rebound in their mental health.

Substance abuse and mental disorders have been defined as the largest contributors to disability for this age group (GBD 2019 Diseases and Injuries Collaborators, 2020), since almost half of the individuals have experienced them by age 29 (Kessler et al., 2005). In fact, most mental health symptoms have their onset in this life stage, although severe disorders are usually diagnosed later (Patel et al., 2007).

* Corresponding author. Department of Psychology and Education Sciences, Universitat Oberta de Catalunya, Rambla del Poblenou 156, 08018, Barcelona, Spain.

E-mail address: [email protected] (J.C. Medina).

Contents lists available at ScienceDirect

Journal of Psychiatric Research

journal homepage: www.elsevier.com/locate/jpsychires

https://doi.org/10.1016/j.jpsychires.2022.06.034

Received 9 March 2022; Received in revised form 26 May 2022; Accepted 10 June 2022

(2)

Of all mental health symptoms that appear in young adults, depression is becoming a major concern. Studies show that 60–80% of youngsters report mild levels of depressive symptoms, and around 5–12% reach severe levels (Schubert et al., 2017). This is unsurprising considering that depression is a debilitating disorder which has become one of the most prevalent (GBD 2019 Diseases and Injuries Collabora- tors, 2020) and a leading cause of burden around the world (Ferrari et al., 2013; Whiteford et al., 2015).

In effect, depression can be experienced at different levels of severity (mild, moderate, or severe) based on the number and intensity of symptoms, and the degree of functional impairment they involve (American Psychiatric Association, 2013). The severity of depression is an important factor clinicians consider when choosing the best treatment for each person. Indeed, several studies (typically those using pharmacotherapy) have reported severity as a significant predictor of treatment efficacy (Fournier et al., 2010; Kirsch et al., 2008). These studies have identified that clinically significant improvement is often only achieved in individuals presenting severe symptoms of depression at the beginning of treatment. In turn, although the literature on intervention efficacy for mild and moderate depression is more scarce, it suggests that psychotherapy, or the combination of psychotherapy and antidepressants, could be favorable in the reduction of symptoms as well (Aherne et al., 2017; Khan et al., 2012). For example, in a study presenting a series of meta-analyses, Cuijpers et al. (2011) found that the efficacy of psychotherapy for mild-to-moderate depression in adults of all ages, not comparing it between age groups, was similar to the efficacy of pharmacotherapy. More specifically, several types of psychological treatment were found to be equally efficacious, including cognitive behavioral therapy, behavioral activation therapy, problem-solving therapy, interpersonal psychotherapy, and non-directive supportive therapy. Recently, psychological Internet- and mobile-based interventions have also proved to be effective in the treatment of depression, and they are considered as promising strategies to provide care to a wider number of people (Josephine et al., 2017).

Although all these studies indicate that psychotherapy is an adequate treatment for mild and moderate depression, it is still not clear yet if their results are also applicable to young people. For example, youths have been proposed to present a differential profile dominated by so- matic symptoms instead of more cognitive or emotional disturbances (Rice et al., 2019), while they also deal with specific challenges that can lead to the presence of depressive symptoms, mostly mild-to-moderate, but more knowledge is needed to understand whether the interventions offered are beneficial to this specific population. Recently, a meta-analysis comparing psychotherapy effects on depression outcomes across different age groups indicated that effect sizes (ESs) were significantly larger in young adults than in middle-aged adults (Cuijpers et al., 2020), but this benefit has not been studied specifically for mild-to-moderate depressive symptoms.

Consequently, the objective of the present study is to conduct a meta- analysis to ascertain whether psychological interventions are efficacious to relieve mild-to-moderate depressive symptoms in young adults. We focus on this population for its preventive value, since several studies demonstrate that this is a stage for the symptoms of most formal mental disorders to appear (Gustavson et al., 2018; Kessler et al., 2005).

Therefore, finding the best early strategies to manage mild-to-moderate depressive symptoms might preclude more severe and chronic courses later in life (Beames et al., 2021; Colizzi et al., 2020).

We expect to retrieve data to compare their efficacy in contrast with no treatment, wait-list, treatment as usual, passive interventions, and other bona-fide treatments. In addition, we are also interested in determining potential moderators of such efficacy (i.e., under which circumstances larger effects are obtained) in order to inform clinical practice in this specific population. Differential efficacy between treatment modalities is one of the paradigmatic variables to be analyzed, but other relevant characteristics of these treatments (e.g., intensity, therapist involvement, delivery means), participants (e.g., age, gender,

severity of depression), and study designs (e.g., quality, geographical region, year of publication) will also be explored, in line with findings regarding potential candidates (e.g., Courtney et al., 2022).

2. Material and methods

2.1. Search strategy and eligibility criteria

This meta-analysis was conducted and is reported according to the meta-analysis reporting standards (MARS; Appelbaum et al., 2018), and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA; Page et al., 2021). The protocol was registered in PROSPERO on April 22, 2020. It was accepted on July 5, 2020 [CRD42020181505].

The search was conducted in Scopus, MEDLINE and PsycINFO; and complemented by a grey literature search in ClinicalTrials.gov, the BMC ISRCTN Registry, Cochrane CENTRAL, Clarivate BIOSIS Previews, and the METAPSY database. The following criteria were used, conducting the search both in English and Spanish: (psychotherap* OR intervention* OR therap* OR treatment*) AND (“mild-to-moderate depress*” OR

“mild to moderate depress*” OR “mild depress*” OR “moderate depress*” OR “subsyndromal depress*” OR “subclinical depress*”) AND (“emerging adult*” OR young* OR youth* OR juvenile OR “16–30 year*” OR “16–30 year*” OR “college student*” OR “university student*” OR “undergraduate student*“ OR “graduate student*“ OR

“postgraduate student*“ OR “post-graduate student*”) AND (efficac* OR effective*).

Dates covered from the start of records to April 27, 2020. Whenever a previous systematic review or meta-analysis was found, citation tracking was performed. Inclusion criteria were: (1) young adult samples (16–30 years old); (2) participants experiencing mild or moderate depressive symptoms; (3) randomized controlled trials (RCTs), non-RCTs, or single- group pre-post studies including a psychological intervention in at least one arm; and (4) administering standardized psychometric measures to assess depressive symptoms. In turn, exclusion criteria were: (1) studies strictly focused on specific medical or psychological conditions other than depression (e.g., pregnancy, epilepsy, chronic pain, cancer, psy- chosis, substance abuse); and (2) qualitative-only studies.

2.2. Review procedure and data extraction

Two independent researchers conducted a blind screening of all documents after removal of duplicates using Rayyan (https://rayyan.

qcri.org/). They followed a stepwise procedure, reviewing the title and abstract in a first stage, and then the full-text of those documents retained in a second phase. After both steps, the reviewers met to discuss disagreements, making final decisions by consensus. The Kappa index after blind review was computed using the “irr" package (Gamer et al., 2019) for “R” software (R Core Team, 2020).

The following data were extracted from all articles: year of publication; study design; mood severity of participants at baseline; mea- surement instruments applied; means and standard deviations; number of participants; interventions administered, their format and intensity;

control group characteristics (if any); gender distribution and age of the samples; and study quality. Data not reported were requested from authors by email.

Finally, the Quality Assessment Tool for Quantitative Studies (QATQS) was used (National Collaborating Centre for Methods and Tools, 2008). This tool covers six areas: selection bias, methodological design, confounders, blinding, data collection methods, and with- drawals and dropouts. Every area is rated (1 =strong, 2 =moderate, 3

=weak) and an average quality score is computed for each document.

2.3. Statistical analyses

Analyses were conducted using the “Meta” package (Schwarzer,

(3)

2007) for “R” software. Random effects were chosen assuming the probability of variation of treatment effects across studies. Reporting bias was assessed using contour-enhanced funnel plots and Egger’s test (Egger et al., 1997). Standardized mean differences (SMD) with 95%

confidence intervals (CIs) were calculated, reporting ESs as Hedges’ g.

To assess heterogeneity among studies we used the I²statistic with 95%

CIs (0% corresponds to no observed, 25% to low, 50% to moderate and 75% to high heterogeneity).

For those studies with more than two groups, we conducted separate analyses for each inter-group comparison, while for studies reporting more than one scale to measure depressive symptoms, their metrics were pooled using Borenstein and colleagues’ (2009) formula. The only exception were studies that measured depressive symptoms from several sources (e.g., self- and clinician-rated instruments). In this case, each

measure was retained since they represented different respondents.

Meta-analytic models were performed to study possible causes of heterogeneity, using tests for subgroup differences and meta- regressions. The former were applied for categorical variables, including baseline severity of depressive symptoms (e.g., mild, moderate); instrument respondent (e.g., self-report, parent-rated); type of psychological treatment (e.g., CBT, narrative), format (e.g., individual, group) and delivery (e.g., face-to-face, online); therapist role (degree of involvement; e.g., classic, minimal guidance); type of control condition (e.g., medication, other psychotherapies); student sample (i.e., yes or no); and region. In turn, meta-regressions were used for continuous variables including year of publication, intervention intensity (in total minutes), participants’ gender balance, participants’ mean age, and study quality (i.e., QATQS). We did not analyze subgroups with less than

Fig. 1.Flow diagram of reviewed studies.

(4)

three studies given their lack of representativeness. Finally, sensitivity analyses were conducted excluding studies with weak quality according to the QATQS to test the robustness of our findings.

3. Results

3.1. Study selection and characteristics

The search retrieved 170 citations in Scopus, 135 in MEDLINE, and 114 in PsycINFO, yielding 419 primary literature documents (see Fig. 1).

Regarding gray literature, we found three studies in ClinicalTrials.gov, one in the BMC ISRCTN Registry, 133 in Cochrane CENTRAL, 32 in Clarivate BIOSIS Previews, and 19 in the METAPSY Database, giving an additional 188 records. There was considerable overlap between these sources, with 294 duplicates removed. Among the 313 documents entering the first-stage review, 207 did not meet inclusion criteria and were excluded, with reviewers reaching a substantial agreement (k = 0.753, p <.001). Among the 106 documents retained after this first step, six were systematic reviews. Their citations were tracked and they were then discarded, adding 17 new articles to the second step. Therefore, a total of 117 articles were reviewed in full text. Among them, 45 finally met all eligibility criteria and were retained for analysis. Again, the agreement between reviewers was substantial (k =0.624, p <.001).

Included studies involved a total of 3,947 individuals (2,132 in intervention and 1,815 in control groups) and were conducted across 15 countries. The quality of these trials was heterogeneous according to the QATQS (see Supplementary Table 1 for specific information about each study).

3.2. Reporting bias

A funnel plot with the SMDs of the selected studies was built (see Fig. 2). Analyses revealed asymmetry for the results (t = − 5.12, p <

.001), this suggesting the presence of a possible reporting bias favoring intervention efficacy. This result remained similar in the sensitivity analysis excluding weak-quality studies and also excluding the two outliers with SMD under − 5.

3.3. Effect of psychological interventions on depressive symptoms and moderator analyses

Results are reported according to study design and comparison

group, given the expected differences in ESs between them, and in alignment with methodological recommendations to classify control groups in psychotherapy trials (Guidi et al., 2018).

3.3.1. Pre-post studies

The three single-group pre-post studies showed an ES for psychological interventions of g = − 0.99 (95% CI = − 1.32, − 0.66, p <.001), with homogeneous results between them (I²=0%; 95% CI =0%, 90%).

Considering the low number of studies of this type, no meta-analytic models could be performed while meta-regressions did not find any significant correlation. Also, all three obtained a QATQS score of 3 (weak quality), therefore sensitivity analyses could not be performed for this category.

The remaining 42 studies were RCTs that used different comparators as control group. Some of them are included in more than one section below because they featured more than two arms.

3.3.2. Psychological intervention vs no treatment

Those that tested psychological interventions against no treatment (n =17), showed an overall ES of g = − 0.68 (95% CI = − 0.87, − 0.48, p

<.001), and this time moderate heterogeneity was found in their results (I²=67%; 95% CI =52%, 78%). Meta-analytic models uncovered significant differences in intervention efficacy according to depressive symptom severity at baseline (Q =7.94, p =.019), with an ES of g =

− 0.40 (95% CI = − 0.60, − 0.19) and low-to-moderate heterogeneity (I²

=45%; 95% CI =0%, 74%) in samples with mild depressive symptoms, g = − 0.91 (95% CI = − 1.25, − 0.57) and moderate heterogeneity (I²= 71%; 95% CI =50%, 83%) in samples with mild-to-moderate depressive symptoms, and g = − 0.93 (95% CI = − 1.57, − 0.30) and moderate heterogeneity (I²=69%; 95% CI =21%, 88%) in samples with moderate depressive symptoms. As can be seen, heterogeneity only decreased in the mild subgroup. The intervention format also showed differences (Q

=6.51, p =.011), with an ES of g = − 0.50 (95% CI = − 0.69, − 0.31) and moderate heterogeneity (I²=59%; 95% CI =29%, 76%) in studies where the intervention was delivered in group format, and up to g =

− 1.16 (95% CI = − 1.63, − 0.69) and moderate heterogeneity again I²= 71%; 95% CI = 47%, 84%) in studies performing individual interventions. Again, heterogeneity remained high. Similarly, the therapist role (see Fig. 3) also proved to be significant (Q =5.20, p =.023), since those studies with classic therapist guidance (i.e., delivering the whole intervention) obtained an ES of g = − 0.73 (95% CI = − 0.95,

− 0.52) and moderate heterogeneity (I²=69%; 95% CI =53%, 79%),

Fig. 2. Funnel plot of the efficacy of psychotherapy on depressive symptoms at post-intervention according to the included studies.

(5)

while those with minimum therapist intervention retrieved an ES of g =

− 0.29 (95% CI = − 0.60, 0.03) and minimum heterogeneity (I²=10%;

95% CI =0%, 91%). This time, it can be noticed that heterogeneity largely improved in the latter but not in the former subgroup. Finally, the efficacy of psychological interventions also depended on study quality (QM =9.25, p =.002), with greater improvements in studies with worse QATQS scores.

Sensitivity analyses revealed a similar overall ES of g = − 0.52 (95%

CI = − 0.70, − 0.34, p <.001) with 14 studies analyzed. Heterogeneity improved slightly but remained moderate (I²=58%; 95% CI =32%, 75%). However, the influence of all the aforementioned variables van- ished in meta-analytic models, with only instrument respondent emerging as significant (Q =4.66, p =.031), with an ES of g = − 0.61 (95% CI = − 0.81, − 0.34) and moderate heterogeneity (I²=58%; 95%

CI =27%, 76%) for self-reported instruments and g = − 0.26 (95% CI =

− 0.49, − 0.03) and minimum heterogeneity (I²=16%; 95% CI =0%, 87%) for clinician-rated measures.

3.3.3. Psychological intervention vs wait-list

Up to 15 studies used wait-list comparison groups, finding an ES of g

= − 1.04 (95% CI = − 1.25, − 0.82, p < .001), although again with moderate-to-high heterogeneity (I²=74%; 95% CI =64%, 81%). Meta- analytic models were performed and found significant differences in the efficacy of psychological interventions depending on baseline depressive symptom severity (Q =12.01, p <.001), with an ES of g = − 0.65 (95%

CI = − 0.84, − 0.46) and relatively homogeneous results (I²=21%; 95%

CI =0%, 62%) in samples with mild-moderate depressive symptoms, and g = − 1.26 (95% CI = − 1.55, − 0.97) with high heterogeneity (I²= 77%; 95% CI = 66%, 84%) in samples with moderate depressive symptoms. These analyses improved heterogeneity in the first but not in the second subgroup. Since only one study had mild severity (Moldovan et al., 2013), it was included in the first subgroup. The intervention delivery also influenced efficacy levels (Q = 4.16, p = .042), with face-to-face interventions achieving an ES of g = − 1.22 (95% CI =

− 1.54, − 0.90) with high heterogeneity (I²=78%; 95% CI =68%, 85%), and online/phone interventions an ES of g = − 0.80 (95% CI = − 1.05,

− 0.56) with moderate heterogeneity (I²=60%; 95% CI =20%, 80%).

However, this variable overlapped with the therapist role since all face-to-face interventions had classic therapist involvement while the online/phone treatments had minimal guidance from professionals.

Results were also different according to whether the samples were composed strictly by students (Q =6.14, p =.013), with an ES of g =

− 1.39 (95% CI = − 1.83, − 0.96) and high heterogeneity (I²=80%; 95%

CI =70%, 87%) for students and g = − 0.80 (95% CI = − 0.99, − 0.61) with moderate heterogeneity (I²=56%; 95% CI =21%, 75%) for varied samples. As can be seen, heterogeneity was still high. Finally, participants’ gender was also significant (QM =7.15, p =.008), since smaller effects were observed in samples with larger proportion of women (see Fig. 4 below).

Performed with 12 studies, sensitivity analyses showed an overall ES Fig. 3. Forest plot of the efficacy of psychotherapy on depressive symptoms at post-intervention compared to no treatment and according to the therapist level of involvement.

(6)

of g = − 1.08 (95% CI = − 1.31, − 0.84, p <.001), while heterogeneity remained between moderate and high (I²=74; 95% CI =63%, 82%), results were close to those obtained from all studies. This time, meta- analytic models uncovered the same moderators as significant, with also similar ES and heterogeneities.

3.3.4. Psychological intervention vs treatment as usual

Only three studies compared psychological interventions against treatment as usual (i.e., referral to the usual interventions available in participants’ community). They showed an ES of g = − 0.12 (95% CI =

− 0.51, 0.26, p =.532), with moderate heterogeneity between studies (I²

=66%; 95% CI =0%, 90%). Given the small number of studies of this type, meta-analytic models could not be computed while meta- regressions found no significant correlations. Sensitivity analyses were not needed in this category since all studies were rated high in quality according to the QATQS.

3.3.5. Psychological intervention vs passive interventions

Five studies used passive interventions as comparator for psychological treatments, retrieving an ES of g = − 0.15 (95% CI = − 0.35, 0.06, p =.155) despite the moderate heterogeneity present in the analyses (I²

= 62%; 95% CI = 31%, 79%). These interventions comprised either supportive contact, self-monitoring or placebo. Again, meta-analytic models could not be performed with only five studies, while meta- regressions uncovered no significant influences. Additional, sensitivity analyses obtained an ES of g = − 0.05 (95% CI = − 0.26, 0.15, p =.597) with only three studies included, and did not improve heterogeneity (I²

=66%; 95% CI =28%, 84%). Again, meta-regressions did not highlight any influence from the variables studied.

3.3.6. Psychological intervention vs other bona-fide interventions Finally, 20 studies compared psychological treatments with other bona-fide interventions (i.e., treatments which were therapeutically intended and based on a clear rationale) and found a global ES of g =

− 0.16 (95% CI = − 0.25, − 0.04, p =.010), but again moderate heterogeneity was found in these results (I²=55%; 95% CI =37%, 68%).

Most studies featured CBT as their experimental group, and the remaining psychotherapeutic approaches were all included in a newly created category called “Others” to allow the estimation of meta- analytic models, embracing cultural values, expressive, gestalt,

mindfulness, narrative, and psychodrama interventions. Four studies compared CBT and this Others category but suggested no significant differences through an ES of g =0.01 (95% CI = − 0.15, 0.17, p =.878), with homogeneity between studies (I²=0%; 95% CI =0%, 60%). In turn, up to 15 studies compared different forms of CBT between them, and two compared interventions included in Others. Such comparisons within the same category were not analyzed because they were deemed uninformative. Differently, the category CBT vs Medication, which only retained two studies (Miranda et al., 2003; Sreevani et al., 2013), could not be analyzed given its lack of representativeness. No influences were found by meta-analytic models and meta-regressions.

Finally, sensitivity analyses with 15 studies revealed a similar overall ES of g = − 0.14 (95% CI = − 0.26, − 0.06, p =.019), with heterogeneity still low-to-moderate (I²=45%; 95% CI =17%, 63%). Three studies were retained in the comparison between CBT and Others, leading to an ES of g =0.01 (95% CI = − 0.17, 0.20, p =.875) and again with homogeneous results (I²=0%; 95% CI =0%, 65%). As it happened for all studies, no significant moderators were found.

4. Discussion

The aim of the present study was to determine the efficacy of psychological interventions for reducing mild-to-moderate depressive symptoms in young adults (16–30 years old). The results were presented according to the study design and the type of comparator.

An overall large ES (g = − 0.99) was identified in pre-post studies, indicating that depressive symptoms improved from baseline to follow up. In addition, results in this category were rather homogeneous.

However, only three studies were analyzed and all of them employed CBT. Consequently, more detailed analyses could not be conducted. The ES was larger than in most of the subsequent categories, but this finding should be cautiously interpreted as it may be influenced by several sources of bias, including those intrinsically linked to the study design (Cuijpers et al., 2016).

In turn, the ES found when contrasting psychological interventions against comparators in RCTs varied from minimal (g = − 0.12 to − 0.16) in studies with active control conditions (i.e., treatment as usual, passive, and bona-fide interventions) to moderate and large when compared to no treatment (g = − 0.68) and wait-list (g = − 1.04) respectively. Thus, it can be concluded that psychological interventions Fig. 4.Bubble plot of the efficacy of psychotherapy on depressive symptoms at post-intervention compared to wait-list and according to the proportion of women in the sample.

(7)

for mild and moderate depressive symptoms are more efficacious than the absence of treatment, but that their effects are only slightly (and not significantly) greater than those of passive interventions and treatment as usual. In general, the results are similar to those found by Cuijpers et al. (2020) for young adults (g =0.98) when meta-analyzing studies comparing psychotherapy with control conditions for depression across age groups. Interestingly, in that study the ES was lower for middle-aged adults (g =0.77). Our findings also resemble those previously reported by Eckshtain et al. (2020) in a recent meta-analysis on the effect of psychotherapy for depressive symptoms on children and adolescents, highlighting a similar response to treatment from childhood to young adulthood. Therefore, better outcomes seem to be expectable for young people than for adults. It could be hypothesized that depressive symptomatology in youngsters is, in general terms, less crystallized than among adults. This may be related to greater cognitive flexibility and the experience of a greater number of vital changes, which can favor symptomatic recovery (Colizzi et al., 2020).

Potential moderators could only be analyzed for studies comparing psychological interventions with no treatment, wait-list, and other bona- fide interventions. For the latter, no significant candidates were found.

On the contrary, in studies including no treatment and wait-list as comparators, the baseline severity of depressive symptoms influenced the treatment efficacy: more severity led to larger effects. Clearly, in severe cases there is more room for improvement than in mild cases, while they are also more in need of treatment, and therefore similar findings have previously been found (Fournier et al., 2010; Kirsch et al., 2008). In consequence, the fact that we have only focused on mild and moderate depressive symptoms is likely to show more modest results for psychological interventions, and may partly account for their non-superiority against passive or usual care. However, we must highlight that this variable was not significant in the sensitivity analysis performed for the studies comparing psychological interventions and no treatment, preventing us from drawing solid conclusions about its relevance.

The therapist role was another significant moderator, with classic involvement (vs. minimal guidance) leading to larger effects. Although most of the studies are in line with this result, there is also some contrasting evidence. For instance, findings from a previous meta-analysis (Cuijpers et al., 2010a) suggested a lack of differences in ESs for treatments in which the therapist had an active role versus interventions in which the therapist provided minimal guidance to support self-help for depression and anxiety. The non-significant status acquired by this variable in the sensitivity analysis for the comparison against no treatment seems to move in such direction. All in all, more research is needed to clarify when and for whom the effect of the therapist role is more relevant, as well as to ascertain if this variable is influential in treatment efficacy by itself or it overlaps with intervention delivery means (i.e., face-to-face or online).

The quality of the studies was a moderator that influenced outcomes only when comparing psychological interventions with no treatment. In these studies, lower quality led to larger effects. These results align with the idea that the effects of psychological treatments for adult depression might be generally overestimated, as highlighted by Cuijpers et al., (2010b). They suggested that such bias may be related to the lack of control of nonspecific factors that contribute to the beneficial effects of treatment in the lower-quality studies (e.g., therapist competence, treatment adherence). In any case, this finding was not supported by the sensitivity analysis for this group of studies, while we must also consider that results in general varied only slightly when excluding weak-quality studies from analyses. Therefore, we cannot underline the relevance of this variable, at least as measured with the QATQS.

Similarly, for those studies comparing psychological interventions with no treatment, individual therapy led to larger effects than group therapy. Cuijpers’ team (Cuijpers et al., 2008) found similar results in a meta-analysis of 15 studies testing psychological treatments for depression in general. However, the study of McDermut and

collaborators (2001) found no statistically significant differences when comparing both formats, and we faced the same scenario of non-significance when leaving the weak-quality studies out in the sensitivity analyses. Therefore, again more research is needed to empirically compare the effects in the mild-to-moderate range of depression of both treatment modalities. At this point, both appear as reasonable options and their differential features (e.g., costs, accessibility) may favor the use of one over the other.

Considering only the studies that compared psychological interventions against wait-list, two sociodemographic moderators were identified: whether the patients were students and their gender distribution. Treatments provided to students resulted in larger effects compared to those provided to varied populations. It can be argued that students have characteristics that might facilitate better outcomes (e.g., higher levels of cognitive functioning, a more stable environment, financial support), explaining the larger effects found. In a previous meta-analysis, a similar hypothesis was proposed (Cuijpers et al., 2016).

However, they compared the effect of treatment in student samples against other depressed adults and found no significant differences.

Indeed, it could also be hypothesized that the greater effects we observed among students may be related to causes other than the intervention; for instance, the credits received in compensation to their participation or their willingness to collaborate with researchers that may also be their lecturers. Also, a smaller proportion of women participating in the studies led to larger effects. One possible reason for this result may be that young adult women in no-treatment conditions are more active than men in trying to engage in help-seeking behaviors (e.g., social support), thus reducing symptoms (Martínez-Hern´aez et al., 2016). An alternative hypothesis is that some vulnerabilities and stressors (e.g., trauma exposure, sexual harassment, limited job oppor- tunities) are more present among women, which might limit treatment response (Hyde and Mezulis, 2020).

Finally, only when performing the sensitivity analysis for the psychological interventions against no treatment subgroup, the assessment respondent was found significant, with larger effects obtained by self- reported measures than for clinician-rated instruments. This result is not aligned with extant literature considering all ages (Cuijpers et al., 2010), but it is when approaching youths (Eckshtain et al., 2020), therefore supporting the hypothesis that they tend to estimate greater improvements in themselves than others do.

The present study provided useful information to understand the efficacy of psychological interventions for subthreshold depression in young adults, although some limitations of the evidence reviewed should be highlighted. The quality of the trials was diverse and less than 50% of the studies were awarded strong quality ratings. This indicates that our results should be interpreted cautiously, and efforts should be directed to improve the quality of future evidence. Also, heterogeneity was a concern in most results, restricting their interpretability. The meta-analytic models proposed tried to adjust the impact of this heterogeneity, however, it remained high in most cases. Another relevant factor to be underlined in the comprehension of the presented findings is the apparent presence of reporting bias, which can be considered robust based on the additional analyses conducted, and may have inflated the effects reported to some extent. We must also note that most of the studies were conducted in North America and Europe, which hinders the generalization of the results to other regions.

Apart from the limitations of the evidence, some constraints related to the review process must also be recognized by authors. To begin with, although we made an effort to cover the most relevant sources of both primary and gray literature, the inclusion of more databases (e.g., Google Scholar) might have complemented our results, especially in the comparisons against usual and passive treatments, which only included three and five studies respectively. Similarly, we were not able to collect information from authors of six studies, resulting in their exclusion from the analyses. Also, although the agreement between reviewers was substantial, it could have been higher especially in the second stage.

(8)

Finally, our findings are limited to the short-term outcomes of treatments (i.e., post-intervention). Future systematic reviews and meta- analyses are invited to focus on longer-term outcomes to study whether psychological interventions in young adulthood prevent depression and distress in later life stages (Beames et al., 2021). They are also suggested to conduct their searches in more languages, since only 31 documents were found in Spanish and all other were published in English.

Despite these limitations, the findings of the present meta-analysis uncover several implications for clinical practice and research. First, the estimated intervention efficacy may serve as a benchmark for future studies and intervention programs tailored to mild-to-moderate depression in young adults. In the light of our results, such programs are expected to significantly reduce depressive symptoms in this population, even more than among adults, and therefore their adoption is encouraged. Also, some variables have been identified as potential moderators of psychological treatment efficacy, like baseline severity and therapist involvement, suggesting that patients with moderate rather than mild symptoms, and receiving treatment by a therapist rather than only minimal guidance, benefited more from therapy. In turn, intervention delivery was only a significant moderator in studies including waitlists. Consequently, it seems that online or face-to-face interventions could be offered to this population, leaving the decision open to considering accessibility and economic factors. The same applies to treatment format (e.g., individual, group), given this variable has only showed significance when psychological interventions were compared to no treatment and did not remain significant in the sensitivity analysis.

Finally, it is equally important to highlight the relevance for practice of those variables that have not proven to be influential. These include the specific type of therapy, treatment intensity (in total minutes) and participants’ age. Therefore, it is relevant to underline that, at this point, CBT is the theoretical approach with more studies published supporting its efficacy for mild-to-moderate depressive symptoms in young adults.

However, such efficacy is likely to be comparable the other psychological interventions, as it has been proven for depression in general (Cuijpers et al., 2021), while all therapies seem to be equally efficacious regardless of their length. Also, specific age does not seem to be influential within the range covered, although it was certainly narrow (i.e., 16–30) given the focus of this meta-analysis.

These results suggest that diverse and equally efficacious pathways to treat mild-to-moderate depressive symptoms in young adults are available, and invite clinicians to contemplate flexible approaches, including but not limited to different delivery means and treatment formats, assessment respondents, theoretical evidence-based approaches, and intervention lengths. Similarly, patients’ age within young adulthood does not seem to play a role in treatment efficacy, while gender and their status as students may do. Consequently, idio- graphic and personalized approaches are supported, in line with previous recommendations (Weisz et al., 2021), and extant literature about the general lack of clearly significant moderators of treatment outcomes (Courtney et al., 2022). Indeed, in most of our results, the 95% CI of effects overlapped, while sensitivity analyses questioned some of the moderators, which should lead to a cautious interpretation. Finally, we encourage researchers from regions other than Europe and North-America to engage in studies to assess the efficacy of psychological interventions for depression in this crucial life period, to gauge the importance of the role of societal, economic and cultural factors.

Funding

This work was supported by the Spanish Ministry of Science, Inno- vation and Universities/Spanish State Research Agency (grant number RTI2018-094294-B-I00); and the Government of Catalonia (grant number 2017SGR642). These funding sources had no influence over the conduction of the research or the results presented herein.

Credit author contribution statement

Joan C. Medina: Conceptualization, Methodology, Formal analysis, Writing - Original Draft, Project administration. Clara Paz: Conceptu- alization, Methodology, Writing - Original Draft, Project administration.

Helena García-Mieres: Investigation, Writing - Review & Editing.

Noelia Nino-Robles: Investigation, Writing - Review ˜ & Editing. Jona- than Herrera: Data Curation. Guillem Feixas: Writing - Review &

Editing, Supervision, Funding acquisition. Adri´an Montesano:

Conceptualization, Writing - Review & Editing, Supervision, Funding acquisition.

Declaration of competing interest None.

Appendix A. Supplementary data

Supplementary data related to this article can be found at https ://doi.org/10.1016/j.jpsychires.2022.06.034.

References

Aherne, D., Fitzgerald, A., Aherne, C., Fitzgerald, N., Slattery, M., Whelan, N., 2017.

Evidence for the treatment of moderate depression: a systematic review. Ir. J.

Psychol. Med. https://doi.org/10.1017/ipm.2017.10.

Appelbaum, M., Cooper, H., Kline, R.B., Mayo-Wilson, E., Nezu, A.M., Rao, S.M., 2018.

Journal article reporting standards for quantitative research in psychology: the APA Publications and Communications Board task force report. Am. Psychol. 73, 3–25.

https://doi.org/10.1037/amp0000191.

Beames, J.R., Kikas, K., Werner-Seidler, A., 2021. Prevention and early intervention of depression in young people: an integrated narrative review of affective awareness and Ecological Momentary Assessment. BMC Psychol 9, 113. https://doi.org/

10.1186/s40359-021-00614-6.

Borenstein, M., Hedges, L.V., Higgins, J.P.T., Rothstein, H.R., 2009. Multiple outcomes or time-points within a study. In: Introduction to Meta-Analysis. John Wiley & Sons, Ltd., Chichester, UK, pp. 225–238.

Colizzi, M., Lasalvia, A., Ruggeri, M., 2020. Prevention and early intervention in youth mental health: is it time for a multidisciplinary and trans-diagnostic model for care?

Int. J. Ment. Health Syst. 14, 23. https://doi.org/10.1186/s13033-020-00356-9.

Courtney, D.B., Watson, P., Krause, K.R., Chan, B.W.C., Bennett, K., Gunlicks- Stoessel, M., Rodak, T., Neprily, K., Zentner, T., Szatmari, P., 2022. Predictors, moderators, and mediators associated with treatment outcome in randomized clinical trials among adolescents with depression. JAMA Netw. Open 5, e2146331.

https://doi.org/10.1001/jamanetworkopen.2021.46331.

Cuijpers, P., Van Straten, A., Warmerdam, L., 2008. Are individual and group treatments equally effective in the treatment of depression in adults? A meta-analysis. Eur. J.

Psychiatry 22, 38–51. https://doi.org/10.4321/S0213-61632008000100005.

Cuijpers, P., Li, J., Hofmann, S.G., Andersson, G., 2010. Self-reported versus clinician- rated symptoms of depression as outcome measures in psychotherapy research on depression: a meta-analysis. Clin. Psychol. Rev. 30, 768–778. https://doi.org/

10.1016/j.cpr.2010.06.001.

Cuijpers, P., Donker, T., van Straten, A., Li, J., Andersson, G., 2010a. Is guided self-help as effective as face-to-face psychotherapy for depression and anxiety disorders? A systematic review and meta-analysis of comparative outcome studies. Psychol. Med.

40, 1943–1957. https://doi.org/10.1017/S0033291710000772.

Cuijpers, P., Van Straten, A., Bohlmeijer, E., Hollon, S.D., Andersson, G., 2010b. The effects of psychotherapy for adult depression are overestimated: a meta-analysis of study quality and effect size. Psychol. Med. 40, 211–223. https://doi.org/10.1017/

S0033291709006114.

Cuijpers, P., Andersson, G., Donker, T., van Straten, A., 2011. Psychological treatment of depression: results of a series of meta-analyses. Nord. J. Psychiatry 65, 354–364.

https://doi.org/10.3109/08039488.2011.596570.

Cuijpers, P., Cristea, I.A., Ebert, D.D., Koot, H.M., Auerbach, R.P., Bruffaerts, R., Kessler, R.C., 2016. Psychological treatment of depression in college students: a metananalysis. Depress. Anxiety 33, 400–414. https://doi.org/10.1002/da.22461.

Cuijpers, P., Karyotaki, E., Eckshtain, D., Ng, M.Y., Corteselli, K.A., Noma, H., Quero, S., Weisz, J.R., 2020. Psychotherapy for depression across different age groups: a systematic review and meta-analysis. JAMA Psychiatry 77, 694. https://doi.org/

10.1001/jamapsychiatry.2020.0164.

Cuijpers, P., Quero, S., Noma, H., Ciharova, M., Miguel, C., Karyotaki, E., Cipriani, A., Cristea, I.A., Furukawa, T.A., 2021. Psychotherapies for depression: a network meta- analysis covering efficacy, acceptability and long-term outcomes of all main treatment types. World Psychiatry 20, 283–293. https://doi.org/10.1002/

wps.20860.

Eckshtain, D., Kuppens, S., Ugueto, A., Ng, M.Y., Vaughn-Coaxum, R., Corteselli, K., Weisz, J.R., 2020. Meta-analysis: 13-year follow-up of psychotherapy effects on youth depression. J. Am. Acad. Child Adolesc. Psychiatry 59, 45–63. https://doi.org/

10.1016/j.jaac.2019.04.002.

(9)

Egger, M., Smith, G.D., Schneider, M., Minder, C., 1997. Bias in meta-analysis detected by a simple, graphical test. BMJ 315, 629–634. https://doi.org/10.1136/

bmj.315.7109.629.

Ferrari, A.J., Charlson, F.J., Norman, R.E., Patten, S.B., Freedman, G., Murray, C.J.L., Vos, T., Whiteford, H.A., 2013. Burden of depressive disorders by country, sex, age, and year: findings from the global burden of disease study 2010. PLoS Med 10, e1001547. https://doi.org/10.1371/journal.pmed.1001547.

Fournier, J.C., DeRubeis, R.J., Hollon, S.D., Dimidjian, S., Amsterdam, J.D., Shelton, R.

C., Fawcett, J., 2010. Antidepressant drug effects and depression severity: a patient- level meta-analysis. JAMA 303, 47–53. https://doi.org/10.1001/jama.2009.1943.

Gamer, M., Lemon, J., Fellows, I., Singh, P., 2019. Irr: Various Coefficients of Interrater Reliability and Agreement.

GBD 2019 Diseases and Injuries Collaborators, 2020. Global burden of 369 diseases and injuries in 204 countries and territories, 1990–2019: a systematic analysis for the Global Burden of Disease Study 2019. Lancet 396, 1204–1222. https://doi.org/

10.1016/S0140-6736(20)30925-9.

Guidi, J., Brakemeier, E.-L., Bockting, C.L.H., Cosci, F., Cuijpers, P., Jarrett, R.B., Linden, M., Marks, I., Peretti, C.S., Rafanelli, C., Rief, W., Schneider, S., Schnyder, U., Sensky, T., Tomba, E., Vazquez, C., Vieta, E., Zipfel, S., Wright, J.H., Fava, G.A., 2018. Methodological recommendations for trials of psychological interventions.

Psychother. Psychosom. 87, 276–284. https://doi.org/10.1159/000490574.

Gustavson, K., Knudsen, A.K., Nesvåg, R., Knudsen, G.P., Vollset, S.E., Reichborn- Kjennerud, T., 2018. Prevalence and stability of mental disorders among young adults: findings from a longitudinal study. BMC Psychiatry 18, 65. https://doi.org/

10.1186/s12888-018-1647-5.

Hyde, J.S., Mezulis, A.H., 2020. Gender differences in depression: biological, affective, cognitive, and sociocultural factors. Harv. Rev. Psychiatry 28, 4–13. https://doi.org/

10.1097/HRP.0000000000000230.

Institute of Medicine and National Research Council of the National Academies, 2015.

Investing in the Health and Well-Being of Young Adults. The National Academies Press. https://doi.org/10.17226/18869.

Josephine, K., Josefine, L., Philipp, D., David, E., Harald, B., 2017. Internet- and mobile- based depression interventions for people with diagnosed depression: a systematic review and meta-analysis. J. Affect. Disord. https://doi.org/10.1016/j.

jad.2017.07.021.

Kessler, R.C., Berglund, P., Demler, O., Jin, R., Merikangas, K.R., Walters, E.E., 2005.

Lifetime prevalence and age-of-onset distributions of DSM-IV disorders in the national comorbidity survey replication. Arch. Gen. Psychiatry 62, 593. https://doi.

org/10.1001/archpsyc.62.6.593.

Khan, A., Faucett, J., Lichtenberg, P., Kirsch, I., Brown, W.A., 2012. A systematic review of comparative efficacy of treatments and controls for depression. PLoS One 7, e41778. https://doi.org/10.1371/journal.pone.0041778.

Kirsch, I., Deacon, B.J., Huedo-Medina, T.B., Scoboria, A., Moore, T.J., Johnson, B.T., 2008. Initial severity and antidepressant benefits: a meta-analysis of data submitted to the food and drug administration. PLoS Med 5, e45. https://doi.org/10.1371/

journal.pmed.0050045.

Martínez-Hern´aez, A., Carceller-Maicas, N., DiGiacomo, S.M., Ariste, S., 2016. Social support and gender differences in coping with depression among emerging adults: a

mixed-methods study. Child Adolesc. Psychiatry Ment. Health 10, 2. https://doi.org/

10.1186/s13034-015-0088-x.

McDermut, W., Miller, I.W., Brown, R.A., 2001. The efficacy of group psychotherapy for depression: a meta-analysis and review of the empirical research. Clin. Psychol. Sci.

Pract. 8, 98–116. https://doi.org/10.1093/CLIPSY.8.1.98.

Miranda, J., Chung, J.Y., Green, B.L., Krupnick, J., Siddique, J., Revicki, D.A., Belin, T., 2003. Treating depression in predominantly low-income young minority women.

Jama 290, 57. https://doi.org/10.1001/jama.290.1.57.

Moldovan, R., Cobeanu, O., David, D., 2013. Cognitive bibliotherapy for mild depressive symptomatology: randomized clinical trial of efficacy and mechanisms of change.

Clin. Psychol. Psychother. 20, 482–493. https://doi.org/10.1002/cpp.1814.

National Collaborating Centre for Methods and Tools, 2008. Quality Assessment Tool for Quantitative Studies. McMaster University, Hamilton, ON.

Page, M.J., McKenzie, J.E., Bossuyt, P.M., Boutron, I., Hoffmann, T.C., Mulrow, C.D., Shamseer, L., Tetzlaff, J.M., Akl, E.A., Brennan, S.E., Chou, R., Glanville, J., Grimshaw, J.M., Hr´objartsson, A., Lalu, M.M., Li, T., Loder, E.W., Mayo-Wilson, E., McDonald, S., McGuinness, L.A., Stewart, L.A., Thomas, J., Tricco, A.C., Welch, V.A., Whiting, P., Moher, D., 2021. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 372, 71. https://doi.org/10.1136/bmj.n71.

Patel, V., Flisher, A.J., Hetrick, S., McGorry, P., 2007. Mental health of young people: a global public-health challenge. Lancet 369, 1302–1313. https://doi.org/10.1016/

S0140-6736(07)60368-7.

Psychiatric Association, American, 2013. Diagnostic and Statistical Manual of Mental Disorders, fifth ed. American Psychiatric Association. https://doi.org/10.1176/appi.

books.9780890425596.

R Core Team, 2020. R: a Language and Environment for Statistical Computing.

Rice, F., Riglin, L., Lomax, T., Souter, E., Potter, R., Smith, D.J., Thapar, A.K., Thapar, A., 2019. Adolescent and adult differences in major depression symptom profiles.

J. Affect. Disord. 243, 175–181. https://doi.org/10.1016/j.jad.2018.09.015.

Schubert, K.O., Clark, S.R., Van, L.K., Collinson, J.L., Baune, B.T., 2017. Depressive symptom trajectories in late adolescence and early adulthood: a systematic review.

Aust. N. Z. J. Psychiatry 51, 477–499. https://doi.org/10.1177/

0004867417700274.

Schwarzer, G., 2007. Meta: an R package for meta-analysis. R News 7, 40–45.

Sreevani, R., Reddemma, K., Chan, C.L.W., Leung, P.P.Y., Wong, V., Chan, C.H.Y., 2013.

Effectiveness of integrated body-mind-spirit group intervention on the well-being of Indian patients with depression: a pilot study. J. Nurs. Res. 21, 178–185. https://doi.

org/10.1097/jnr.0b013e3182a0b041.

The Society for Adolescent Health and Medicine, 2017. Young adult health and well- being: a position statement of the society for adolescent health and medicine.

J. Adolesc. Heal. 60, 758–759. https://doi.org/10.1016/j.jadohealth.2017.03.021.

Weisz, J.R., Fitzpatrick, O.M., Venturo-Conerly, K., Cho, E., 2021. Process-based and principle-guided approaches in youth psychotherapy. World Psychiatry 20, 378–380. https://doi.org/10.1002/wps.20887.

Whiteford, H.A., Ferrari, A.J., Degenhardt, L., Feigin, V., Vos, T., 2015. The global burden of mental, neurological and substance use disorders: an analysis from the global burden of disease study 2010. PLoS One 10, 1–14. https://doi.org/10.1371/

journal.pone.0116820.