General v. specific vulnerabilities: polygenic risk scores and higher-order psychopathology dimensions in the Adolescent Brain Cognitive Development (ABCD) Study

Background Polygenic risk scores (PRSs) capture genetic vulnerability to psychiatric conditions. However, PRSs are often associated with multiple mental health problems in children, complicating their use in research and clinical practice. The current study is the first to systematically test which PRSs associate broadly with all forms of childhood psychopathology, and which PRSs are more specific to one or a handful of forms of psychopathology. Methods The sample consisted of 4717 unrelated children (mean age = 9.92, s.d. = 0.62; 47.1% female; all European ancestry). Psychopathology was conceptualized hierarchically as empirically derived general factor (p-factor) and five specific factors: externalizing, internalizing, neurodevelopmental, somatoform, and detachment. Partial correlations explored associations between psychopathology factors and 22 psychopathology-related PRSs. Regressions tested which level of the psychopathology hierarchy was most strongly associated with each PRS. Results Thirteen PRSs were significantly associated with the general factor, most prominently Chronic Multisite Pain-PRS (r = 0.098), ADHD-PRS (r = 0.079), and Depression-PRS (r = 0.078). After adjusting for the general factor, Depression-PRS, Neuroticism-PRS, PTSD-PRS, Insomnia-PRS, Chronic Back Pain-PRS, and Autism-PRS were not associated with lower order factors. Conversely, several externalizing PRSs, including Adventurousness-PRS and Disinhibition-PRS, remained associated with the externalizing factor (|r| = 0.040–0.058). The ADHD-PRS remained uniquely associated with the neurodevelopmental factor (r = 062). Conclusions PRSs developed to predict vulnerability to emotional difficulties and chronic pain generally captured genetic risk for all forms of childhood psychopathology. PRSs developed to predict vulnerability to externalizing difficulties, e.g. disinhibition, tended to be more specific in predicting behavioral problems. The results may inform translation of existing PRSs to pediatric research and future clinical practice.

Can parent-infant observation predict later childhood psychopathology: a systematic review

AimsDifficulties in parent-child interaction are easily observed and are a potential target for early intervention. This study aimed to assess the utility of observation of parent-child interaction in the first year of life in identifying children at risk of developing later psychopathology, using a rigorous systematic review method.MethodEMBASE, CINAHL, PsycINFO, MIDIRS, MEDLINE and Cochrane Library databases were searched using MeSH terms and keywords, and reference lists screened. Two authors independently reviewed papers for inclusion and completed data extraction. All peer reviewed papers studying the association between an independent observation of parent-child interaction and later childhood psychopathology in community-based samples were included. Studies based on ‘high risk’ samples (studies exclusively examining cohorts with a sibling or parent with a mental illness or studies of low birth weight or premature infants and those with other physical comorbidities) were excluded. Results were synthesised qualitatively due to high heterogeneity.Result18,226 papers were identified, nine were included in this study. Childhood psychopathology was associated with fewer positive parent-infant interactions, lower parent vocalisation frequency and lower levels of adult speech and activity. Maternal sensitivity was inversely related to separation anxiety and oppositional defiant/conduct disorders were associated with lower shared look rates. Disruptive behaviour disorders were associated with higher frequency of child vocalisation. Pervasive developmental disorders were associated with ‘abnormal’ maternal infant interactions, as assessed by community health nurses using a standardised measure.ConclusionIncluded studies reported small samples, and several of these samples overlapped. Some studies were of poor quality, but were included due to a paucity of available data. The findings may therefore have limited generalisability. Difficulties in parent-child interaction are easily observed and assessments could be made by non-specialists such as health visitors or general practitioners. Such difficulties may be an early indicator of later childhood psychopathology. Childhood psychiatric diagnoses (with the exception of Autistic Spectrum Disorders) appear associated with level of maternal activity (vocalisation, physical activity, positive parenting and shared attention). Assessments may identify at-risk families for early intervention, but further work is required to develop and validate reliable methods for risk stratification in community-based practice.

General vs. Specific Vulnerabilities: Polygenic Risk Scores and Higher-Order Psychopathology Dimensions in the Adolescent Brain Cognitive Development (ABCD) Study

Background. Polygenic risk scores (PRS) capture genetic vulnerability to psychiatric conditions and promise to advance our understanding of mental health etiology in children. Emerging evidence suggests that PRSs may be associated with higher-order dimensions of childhood psychopathology. The current study delineated a pattern of associations of major PRSs with an overarching general factor of psychopathology (p-factor), and five specific factors: externalizing, internalizing, neurodevelopmental, somatoform, and detachment.Method. The sample consisted of 4,717 unrelated children (mean age=9.92, SD=.62; 47.1% female; all European ancestry). Psychopathology was conceptualized hierarchically as empirically-derived general factor and five specific factors. Partial correlations explored associations between psychopathology factors and major psychopathology-related PRSs originally discovered in large samples (Ns>100,000). Regressions tested which level of the psychopathology hierarchy was most strongly associated with each PRS.Results. Four PRSs were associated primarily with the general factor (>60% of genetic effects were general): Depression-PRS, Neuroticism-PRS, PTSD-PRS, and Insomnia-PRS. Two PRS contributed comparably to general and specific psychopathology: Smoking-PRS and Number of Sexual Partners-PRS. Five PRSs contributed primary to specific factors (<40% of genetic effects were general): Adventurousness-PRS, Disinhibition-PRS, Educational Attainment-PRSs, BMI-PRS, and Intelligence-PRS. The incremental associations with specific factors were mainly driven by the externalizing dimension. Conclusion. The PRSs for internalizing problems predominantly captured non-specific genetic vulnerability to psychopathology in children. Conversely, PRSs for externalizing problems contributed to more specific psychopathology outcomes, most notably externalizing behavior. Overall, many major PRSs captured both general and specific genetic vulnerability to childhood psychopathology.