Mental health of adolescents associated with sexual and reproductive outcomes: a systematic review

Abstract Objective To systematically review the literature on the mental health of adolescents associated with sexual and reproductive outcomes, and compare the mental health outcomes with that of other age groups. Methods We searched seven databases for relevant peer-reviewed articles published between 1 January 2010 and 25 April 2019. Our inclusion criteria required that the study included age-disaggregated data on adolescents, and focused and assessed mental health outcomes associated with pregnancy or sexually transmitted infections. We extracted data on the specific health event, the mental health outcome and the method of measuring this, and comparisons with other age groups. Findings After initially screening 10 818 articles by title and abstract, we included 96 articles in our review. We observed that a wide-ranging prevalence of mental ill-health has been reported for adolescents. However, most studies of mental health during pregnancy did not identify an increased risk of depression or other mental disorders among adolescents compared with other age groups. In contrast, the majority of studies conducted during the postpartum period identified an increased risk of depression in adolescents compared with other age groups. Three studies reported on mental health outcomes following abortion, with varying results. We found no studies of the effect of sexually transmitted infections on mental health among adolescents. Conclusion We recommend that sexual and reproductive health services should be accessible to adolescents to address their needs and help to prevent any adverse mental health outcomes.


Introduction
In many countries, adolescents (i.e. those aged 10-19 years) struggle to access necessary sexual and reproductive health information and services. 1 Complications during pregnancy and childbirth are the leading cause of death globally for girls aged 15-19 years. 2 One in four sexually active adolescents has a sexually transmitted infection, and 3 million girls aged 15-19 years undergo unsafe abortions annually. 3 Although the effect of a sexual and reproductive health event (e.g. pregnancy or sexually transmitted infection) on an adolescent's physical health and well-being is acknowledged, 1 the global mental health burden that may be related to the outcomes of sexual activity is not well understood. A 2009 review conducted by the World Health Organization (WHO) identified close links between women's sexual and reproductive health and their mental health. 4 However, many of the participants of this review were married women of childbearing age (i.e. often not adolescents) in middle-and high-income countries. The link between the sexual and reproductive health and the mental health of men and young, single women remains largely unexplored. 4 We therefore conducted a systematic review to examine the impact of key sexual and reproductive health events on mental health outcomes among adolescents. Specifically, we focused on events that can occur as a result of unprotected sexual activity, for example: pregnancy; the result of that pregnancy; and/or sexually transmitted infections, including human immunodeficiency virus (HIV). Our systematic re-view addressed two main areas: (i) the adverse mental health outcomes experienced by adolescents worldwide following key sexual and reproductive health events; and (ii) how this mental health burden among adolescents compares with that of people of other ages after experiencing the same event.

Search strategy
Our protocol was adapted from that of a prior systematic review of causes of maternal morbidity and mortality, 5 and was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. We identified related publications by searching the databases PubMed®, CINAHL, Em-base®, APA PsycINFO®, POPLINE, ERIC (Education Resources Information Center) and Global Index Medicus, as well the reference lists of relevant articles. We developed our search strategy for each database in collaboration with a librarian, using terms related to "mental health", "adolescents" and "sexual and reproductive health"; we provide an example search strategy in the data repository. 6

Selection criteria
Our search included peer-reviewed literature published between 1 January 2010 and 25 April 2019. We selected this particular date range to capture the most recent literature, but also to build on the 2009 WHO review of the mental health aspects of women's reproductive health. 4 Our inclusion criteria required that the study: included age-disaggregated data on male and/or female adolescents; fo-Objective To systematically review the literature on the mental health of adolescents associated with sexual and reproductive outcomes, and compare the mental health outcomes with that of other age groups. Methods We searched seven databases for relevant peer-reviewed articles published between 1 January 2010 and 25 April 2019. Our inclusion criteria required that the study included age-disaggregated data on adolescents, and focused and assessed mental health outcomes associated with pregnancy or sexually transmitted infections. We extracted data on the specific health event, the mental health outcome and the method of measuring this, and comparisons with other age groups. Findings After initially screening 10 818 articles by title and abstract, we included 96 articles in our review. We observed that a wide-ranging prevalence of mental ill-health has been reported for adolescents. However, most studies of mental health during pregnancy did not identify an increased risk of depression or other mental disorders among adolescents compared with other age groups. In contrast, the majority of studies conducted during the postpartum period identified an increased risk of depression in adolescents compared with other age groups. Three studies reported on mental health outcomes following abortion, with varying results. We found no studies of the effect of sexually transmitted infections on mental health among adolescents. Conclusion We recommend that sexual and reproductive health services should be accessible to adolescents to address their needs and help to prevent any adverse mental health outcomes.
cused on mental health outcomes associated with either a pregnancy, the result of that pregnancy (either childbirth and the postpartum period, or an abortion) and/or horizontally transmitted sexually transmitted infections; assessed mental health outcomes that followed a sexual and reproductive health event; and was peer-reviewed. Because of the inconsistency in the literature on the exact definition of the postpartum period, 7 we included any articles referencing the postpartum period as within one year following childbirth. In the case of a randomized controlled trial or intervention study, we also included data from the control group.
We excluded studies that: had sample sizes less than 50; did not disaggregate adolescent-only data; did not quantify mental health outcomes; did not clarify that the sexual and reproductive health event preceded the mental health outcome of interest; used a sample group that was not representative of a general, healthy population (e.g. we excluded studies that recruited only individuals: (i) with specific pre-existing conditions such as type 1 diabetes mellitus, or a mental health condition; or (ii) exhibiting specific behaviours, such as injecting drugs); or were published in languages other than English, French, Italian, Portuguese, Spanish or Turkish.
Following the removal of duplicates, we initially screened the articles by title and abstract before screening the remaining articles in full. The two reviewers assessed and categorized articles as include, unsure or exclude, resolving discrepancies through discussion. A third reviewer, whose judgement was considered final, adjudicated unresolved disputes.

Data extraction
We extracted data on the general study characteristics, the specific sexual and reproductive health event and the mental health outcome of interest (e.g. relative risk among adolescents compared with other age groups). We also extracted comparison data for other age groups if it was presented in a way that assessed the statistical difference between age groups, for example, odds ratio (OR) or relative risk with 95% confidence intervals (CIs). The two reviewers who conducted the initial title and abstract screen independently extracted relevant data using an extraction form. All extracted data were double-checked and confirmed by the other reviewer, and the third reviewer resolved disagreement in the same manner as for study inclusion. Given the diversity in the study designs, measurement tools and definitions adopted in the included articles, we did not perform a metaanalysis of the findings; instead, we summarized outcomes according to type of sexual and reproductive health event.
We used a modified Joanna Briggs Institute critical appraisal checklist to assess the quality of studies reporting prevalence data. 8 We assessed papers for quality according to eight criteria (data repository), 6 and assigned each criterion a score of either 0 (not fulfilled), 1 (unclear whether fulfilled) or 2 (fulfilled); possible scores ranged from 0 to 16. We developed categories of quality, and considered scores of ≤ 12, of 13 or 14, and of 15 or 16 to represent studies of low, medium and high quality, respectively. We also included the stipulation that high-quality articles must score 2 points for the fifth criterion, that is, the study used objective, validated criteria to measure the mental health outcome. The same two reviewers conducted separate quality assessments for all articles, with the same third reviewer resolving differences through discussion.

Results
We initially screened 10 818 articles by title and abstract, after removing duplicates; 9559 articles were immediately excluded. Following full-text review, we excluded another 1112 articles. We therefore included 96 articles in our review, spanning 26 different countries (Fig. 1). Forty-eight studies were conducted in high-income countries, 36 in upper-middle-income countries, 10 in lower-middle-income countries and only two in low-income countries. The United States of America was the country most represented with 38 studies, followed by Brazil with 22. Most (55 studies) were of medium quality and around one third (33 studies) were of  Around one half of the articles reported on the prevalence or mean/ median of mental health conditions or symptoms during pregnancy (48 studies) and/or the postpartum period (51 studies). Three articles reported on the impact of abortion. None of our included studies reported on horizontally transmitted infections. All studies included only women in their samples with the exception of two: one study included both males and females 9 and another included males only. 10 We observed that a variety of tools were used to assess for mental health conditions, including symptom assessment scales or diagnostic tools or codes. Box 1 provides a summary of the tools used and the observed frequency of use among our included studies.

Pregnancy
Of the 48 studies focusing on the prevalence or mean/median of mental health conditions or symptoms during pregnancy (Table 1 available  Of the nine studies (four of which also reported on depression) reporting on some type of anxiety disorder or symptoms during pregnancy, six studies provided prevalence, and the prevalence of at least mild symptoms or an anxiety disorder (including post-traumatic stress disorder) ranged from 13.6% 21 to 19.2%. 48 Four studies reported on mean or median anxiety scores. Six studies (three of which also reported on depression and/or anxiety) reported on broad mental disorders or common mental disorders or stress during pregnancy, with prevalence ranging from 22.5% 17 to 40.5%. 53 One study reported an incidence rate of 4.9%. 55 Four studies (one of which also reported on depression, anxiety and general mental disorders) reported on suicidal ideation or behaviour. The prevalence of any suicidal ideation (including thoughts of selfharm or wishes to be dead) ranged from 4.2% 21,47 to 8.9%, 56 while the prevalence of any suicidal behaviour ranged from < 0.1% (147/1 242 318) 57 to 13.3%. 21,47 Eleven studies provided data regarding depression among adolescents compared with other age groups during pregnancy (Table 2). Of these, only three studies identified an increased risk of depression among pregnant adolescents when compared with older age groups; eight studies reported no increased risk. Five studies reported comparison data regarding general mental disorders or psychological distress among adolescents compared with other age groups during pregnancy. Again, the majority of these studies (four studies) did not identify an increased risk of general mental health problems during pregnancy among adolescents compared with other age groups. One study reported on adjusted OR for suicidal ideation during pregnancy, and found adolescents to be at greatest risk compared with other age groups. 56

Postpartum
In the 49 studies that reported on the prevalence or mean/median of mental health conditions during the postpartum period (Table 3; 47 studies; available at: http:// www .who .int/ bulletin/ volumes/ 99/ 5/ 20 -54144) or during both pregnancy and the postpartum period (Table 4; 2 studies), we noted that the postpartum period was defined as being as short as 72 hours to as long as 1 year after delivery. The majority of these studies (46/49) reported on depression, most (38/46) reporting prevalence of depression or depressive symptoms (i.e. having at least mild symptoms at some time during the postpartum period) from 2.5% 88 to 57.0%. 70 Two studies Box 1. Systematic review of adolescent mental health following sexual and reproductive health events, 2020: frequency of mental health assessment tools used in 96 included studies reported on incidence, which was found to be 25.0% (95% CI: 13.2-36.8%) in one study, 74 and 8.3% at 6 weeks, 5.2% at 3 months and 6.2% at 6 months postpartum in the other. 77 Seven studies reported mean or median symptom scores. One study (Table 4) reported on the prevalence of depression during both pregnancy and the postpartum period. 97 Only three studies focusing on the postpartum period alone reported on anxiety (as well as depression; Table 3), one of which reported the prevalence of any anxiety disorder as 15.3%. 61 The other two studies provided results in the form of mean or median scores. 18,44 Two studies reported on psychiatric disorders or psychological distress during the postpartum period (one of which also reported on depression), reporting a prevalence of 22.6% 61 and 10.0%, 95 respectively. Finally, one study reported the prevalence of suicidal risk during this period as 13.8%. 96 Of the 13 studies that determined the risk of depression during the postpartum period in adolescents compared with other age groups, nine studies identified an increased risk of depression for adolescents. A study reporting on postpartum anxiety (as well as depression) did not find adolescents to be at a higher risk than other age groups (Table 5). 101 However, a study examining suicide risk during this period found adolescents to be at the greatest risk of suicide compared with other age groups. 96

Abortion
We list the three included studies on mental health outcomes following an induced abortion among adolescents in Table 6. Two studies reported that the abortion took place within 12 weeks gestation; 103,104 the third study did not specify when the abortion took place. 67 Two studies reported the prevalence of depressive symptoms as 16.1% 67 and 85.0%, 103 reporting at least mild symptoms of depression. One study reported an average depression score 104 and another provided mean anxiety scores. 103 We did not identify any studies that compare mental health outcomes among adolescents with those of other age groups post-abortion.

Systematic reviews
Adolescent mental health following sexual and reproductive health events Rachel Vanderkruik et al.
outcomes among adolescents after key sexual and reproductive health events reveals a very high prevalence of mental ill-health. This finding is particularly true for depression, the most commonly assessed mental health outcome in our review. The prevalence of depression varies widely between cultures; 105 however, the WHO World Health Survey of 60 countries found an average annual prevalence of 3.2% in participants without comorbid physical disease. 106 Global estimates indicate that 10% of pregnant women and 13% of postpartum women (of all ages) experience a mental disorder. 107 However, among studies reporting the prevalence of depression in this review, 92.9% (26/28) of studies of pregnant adolescents, and 73.0% (27/37) of studies of postpartum adolescents, reported a higher figure than these global estimates.
Our results indicate a high prevalence of depression during pregnancy across all age groups, highlighting the importance of recognizing the mental health needs of all women during pregnancy. By contrast, while not conclusive, comparison data for the postpartum period suggests that there may be a higher prevalence of depression among postpartum adolescents than among postpartum women of older age groups. This higher prevalence among adolescents may be the effect of the challenges facing adolescent mothers in caring for an infant, such as financial burden, social isolation from peers, limited support and the detrimental consequences of being excluded from further education. Our findings also corroborate what is known about the adverse effects of early marriage and the often-ensuing expectation to give birth. 108 Regardless of the nature of the sexual and reproductive health event, it appears that suicidality may be a concern for adolescents. Although only two studies provided comparison data on suicidality between age groups, both showed the risk to be significantly higher among adolescents compared with older age groups. 56,96 Such findings are consistent with other research suggesting that adolescent mothers may have an elevated risk of suicidal ideation. 109 This result emphasizes the need for both further research and enhanced services.
Although the assessments of mental health conditions other than depression were relatively limited, we found a high burden of anxiety and common mental  (. . .continued) Adolescent mental health following sexual and reproductive health events Rachel Vanderkruik et al.
health disorders among adolescents following sexual and reproductive health events. We were surprised to discover only a few studies reported on anxiety, 110 given that depression and anxiety are often comorbid and that the prevalence of anxiety is high among adolescents; a study based in the USA found that 31.9% of adolescents have an anxiety disorder. 111 Our findings point to the need for research on the full spectrum of mental disorders to fully understand the mental ill-health burden experienced by adolescents following such health events, concurring with other calls for research on a broader range of mental health conditions among perinatal women of all ages. 112,113 Our findings reveal another gap in the mental health literature; we found only three articles that reported on abortion and zero articles reporting on the prevalence of sexually transmitted infections among adolescents. Our search did identify 46 articles reporting on HIV; however, we had to exclude these from our review because HIV was either vertically transmitted (i.e. not the result of an unprotected sexual event) or the reviewers were unable to distinguish between mental health outcomes for vertically and horizontally transmitted HIV. Furthermore, while our review included studies on both males and females, nearly all of the included studies (with the exception of two) focused on adolescent females. While this finding is understandable for the reproductive events of pregnancy and abortion, there is an obvious need for more research on the mental health of adolescent males as a result of relevant sexual and reproductive health events (e.g. sexually transmitted infection/HIV, new fatherhood).
We also found few studies describing mental health outcomes following an abortion, highlighting another important area for further research. A study 67 found depression rates to be lower among female adolescents with unintended pregnancies who had an abortion, compared with those who delivered, recorded either one year after abortion or delivery. Although this is only one study, this result supports the notion that when afforded the right to choose, women who elect to have an abortion rarely regret it. 114 While our systematic review has several strengths -such as considering literature published over an entire decade, the thoroughness of the search, (ii) their potential to experience a sexual and reproductive health event; and/or (iii) the incidence and severity of any resulting mental health outcomes. While we were interested in mental health outcomes following a sexual and reproductive health event, we could not always be certain about the exact temporal relationship; it is possible that mental health issues may have increased vulnerability to the particular sexual and reproductive health event. We also excluded qualitative studies from this review; although qualitative data can provide a rich understanding of the impact of such health events on the mental health and wellbeing of adolescents, a mixed-methods systematic review was beyond our scope.
Finally, our quality assessment tool did not undergo a formal psychometric evaluation; however, we based our quality assessment tool on an existing and widely used instrument, which was deemed to have content validity and was used by two authors independently. We felt that this instrument was adequate for our objective of providing an explicit indication of study quality, rather than a precise measurement. We identified methodological issues in many of the included studies. Most studies used assessment tools that screen for the severity of symptoms, but cannot provide a mental health diagno-sis. We observed that a broad range of assessment tools were used, as well as different cut-off points for the same tool between different studies. For example, one study 31 used a Center for Epidemiologic Studies Depression scale (CES-D) cut-off of ≥ 16, whereas another study 16 used one of ≥ 24. Almost half of the included studies used the Edinburgh Postnatal Depression Scale (EPDS); of these studies, many (21/40) used a cutoff of ≥ 13 although others used scores of 10 15 or 9. 27 The field of mental health would benefit from the streamlining of screening tools and cut-offs used, which would also encourage research that uses diagnostic tools to confirm mental health conditions rather than solely identifying symptom severity. The use of rigorous clinical diagnostic interviews to assess mental health disorders would provide a clearer clinical picture of the mental health burden among adolescents who have experienced a sexual and reproductive health event.
To address the mental health burden associated with pregnancy or sexually transmitted infections, future work should identify effective psychosocial interventions that can be made available to adolescents who experience such a health event. There is evidence that adolescents often do not use mental health services, so these efforts should consider how to successfully connect identified adolescents with the care they need. 115,116 Additional research could also identify risk and protective factors in adolescents who have experienced such a health event by comparing those who developed mental health issues with those who did not; this strategy may help to determine whether targeted interventions can build resiliency among adolescents who experience such a health event. Similarly, research is also needed to explore the extent to which adolescents experience adverse mental health outcomes, comparing those who previously experienced a sexual and reproductive health event with those who did not. As a promising step, WHO published the Guidelines on mental health promotive and preventive interventions for adolescents in 2020, with recommendations targeting all adolescents and particularly vulnerable groups. 117 In conclusion, considering the mental health burden that adolescents are experiencing, we now need to develop, implement and evaluate appropriate services to support the adolescent population. Similarly, sexual and reproductive health services and information should be accessible to adolescents to address their needs and help to prevent any unintended outcomes that could have consequences for their mental health. Health-care providers encountering adolescents following such a health event must be prepared to screen for, and address, any mental health concerns. Going forwards, we recommend that mental health care is considered an integral part of sexual and reproductive health service provision. ■   constaté une grande diversité de prévalence des troubles mentaux chez les adolescents concernés. La plupart des études s'intéressant à la santé mentale pendant la grossesse n'ont toutefois pas identifié plus de risques de dépression ou d'autres troubles mentaux dans cette tranche d'âge que dans d'autres. En revanche, la majorité des études menées au cours de la période post-partum ont fait état d'un risque accru de dépression chez les adolescents par rapport à d'autres catégories. Trois études mentionnent des problèmes de santé mentale après un avortement, avec divers résultats. Nous n'avons trouvé aucune étude consacrée aux répercussions des infections sexuellement transmissibles sur la santé mentale des adolescents. Conclusion Nous recommandons de rendre les services de santé sexuelle et reproductive accessibles aux adolescents afin de répondre à leurs besoins et de limiter l'impact négatif sur leur santé mentale.

Salud mental de los adolescentes asociada a los resultados sexuales y reproductivos: revisión sistemática
Objetivo Revisar sistemáticamente las publicaciones sobre la salud mental de los adolescentes asociada a los resultados sexuales y reproductivos, y comparar los resultados de salud mental con los de otros grupos de edad. Métodos Se realizaron búsquedas en siete bases de datos de artículos relevantes revisados por expertos y publicados entre el 1 de enero de 2010 y el 25 de abril de 2019. Nuestros criterios de inclusión requerían que el estudio incluyera datos desglosados por edad de los adolescentes, centrados y que evaluaran los resultados de salud mental asociados con el embarazo o las infecciones de transmisión sexual. Se extrajeron datos sobre el evento de salud específico, el resultado de salud mental y el método de medición de este, así como las comparaciones con otros grupos de edad.
Observamos que se ha informado de una amplia prevalencia de mala salud mental en las adolescentes. Sin embargo, la mayoría de los estudios sobre salud mental durante el embarazo no identificaron un mayor riesgo de depresión u otros trastornos mentales entre las adolescentes en comparación con otros grupos de edad. En cambio, la mayoría de los estudios realizados durante el período posparto identificaron un mayor riesgo de depresión en las adolescentes en comparación con otros grupos de edad. Tres estudios informaron sobre los resultados de la salud mental después del aborto, con resultados variados. No se encontraron estudios sobre el efecto de las infecciones de transmisión sexual en la salud mental de las adolescentes. Conclusión Recomendamos que los servicios de salud sexual y reproductiva sean accesibles a los adolescentes para atender sus necesidades y ayudar a prevenir cualquier resultado adverso para la salud mental.