Risk and protective factors for mental disorders with onset in childhood/adolescence: An umbrella review of published meta-analyses of observational longitudinal studies

Highlights

• This is the first umbrella review grading risk factors for disorders with onset during childhood/adolescence.

• No risk factor is supported by convincing evidence nor by highly suggestive evidence.

• Exposure to antipsychotics/lithium during pregnancy as risk factor for neuromotor deficits is supported by suggestive evidence, but confounding by indication is present.

• No protective factor has been identified.

Abstract

The patho-etiology of mental disorders with onset in childhood or adolescence remains largely unknown. We conducted an umbrella review of meta-analyses (MAs) on environmental factors associated with mental disorders with onset in childhood/adolescence. We searched Pubmed-MEDLINE/EMBASE/PsycInfo databases, last search April 29th, 2020. Quality of MAs was measured with AMSTAR-2. Out of 6851 initial references, ten articles met inclusion criteria, providing 23 associations between 12 potential environmental factors and nine disorders (cases: 8884; N = 3,660,670). While almost half of the associations were nominally significant, none of them met criteria from either convincing or highly suggestive evidence. A single association was supported by suggestive evidence (maternal exposure to lithium or antipsychotics with neuromotor deficits), but it was affected by confounding by indication. Ten more associations had weak evidence, and 12 associations were not statistically significant. Quality of meta-analyses was rated as high in two, moderate in one, low in four, critically low in two, and not pertinent in one (individual participant data). Methodologically-sound research is needed in this field.

Introduction

It is commonly recognized that a large portion of mental disorders have their onset before age 18 (Paus et al., 2008). However, some of them have an even earlier onset. Mental disorders with onset in childhood are grouped differently in the 10th and 11th versions of the International Classification of Diseases ((ICD)-10 (WHO, 2019a) and ICD-11 (WHO, 2019b), respectively). In ICD-10, they are classified under “V - Mental and behavioural disorders”, “F80-F89 Disorders of psychological development”, and “F90-F98 Behavioural and emotional disorders with onset usually occurring in childhood and adolescence”, while in the ICD-11 they are grouped under “06 - Mental, behavioural or neurodevelopmental disorders”, “Neurodevelopmental disorders”, and “Disruptive behaviour or dissocial disorders”. Despite being defined as individual categorical entities, or even being clustered in different blocks within and across ICD-10 and ICD-11, the above listed disorders are characterised by their onset, which is typically in childhood. Importantly, in terms of clinical presentation, there is a high degree of overlap among some of these disorders, which provides a strong rationale for considering them together when exploring their putative patho-etiology (Thapar et al., 2017). These childhood-onset disorders present complex and heterogeneous patho-etiology, and, although highly heritable (Lord et al., 2018; Yousaf et al., 2020; Demontis et al., 2019; Lichtenstein et al., 2010), their aetiology is typically multifactorial (with the exception of cases with single major causes, such as patients with fetal alcohol syndrome or specific genetic syndromes). Several putative environmental risk and protective factors have been proposed and assessed in several meta-analyses, with conflicting results. Gaining insight on credible risk or protective factors for neurodevelopmental disorders is a crucial initial step to illuminate the development of preventative strategies for these disorders aimed at reducing the significant financial and societal burden related to mental disorders with onset in childhood (Arim et al., 2017; Arango et al., 2018). While two recent umbrella reviews (Kim et al., 2019, 2020) have graded the available evidence from meta-analyses of observational studies on risk factors for autism-spectrum disorder (ASD) (Kim et al., 2019), and attention-deficit/hyperactivity disorder (ADHD), Kim et al., 2020) respectively, the credibility of the claimed associations between putative risk factors for other neurodevelopmental disorders and disorders with onset in childhood remains unknown. To fill this gap in the literature, we conducted an umbrella review focused on environmental risk and protective factors for disorders with onset in childhood other than ADHD or ASD. Of note, we endeavoured to address possible methodological issues when assessing putative risk factors. Indeed, as shown in several previous umbrella reviews on risk factors for mental disorders or obesity, (Kim et al., 2019; Solmi et al., 2020; Köhler et al., 2018; Bortolato et al., 2017; Solmi et al., 2018a) several sources of bias are present across studies, which individually or altogether can contribute to over- or underestimate the associations between socio-demographic, health, environmental, and other risk factors and disorders. Additionally, when medications are considered as potential risk factors, confounding by indication can affect several false positive claimed associations between medications and disorders’ onset (Solmi et al., 2018b). Therefore, the aim of this umbrella review was to grade the evidence from meta-analyses of cohort and case-control studies on protective and risk factors for the most important mental disorders with onset in childhood accounting for several sources of bias and applying established quantitative criteria.