Genomics in the News

This was the first publication on pandemic-related mental health changes in child psychiatric outpatients in the United States. Empirically-derived multivariate models showed that child psychiatric outpatients had highly variable emotional and behavioral reactions to the pandemic, with some showing stability, others even showing improvement, and two groups manifesting distinctive patterns of worsening symptoms predicted by their prior psychopathology. Results have implications for clinical care as well as ongoing research on genetic contributions to youth mental health.

Polygenic risk scores (PRS) have attenuated cross-population predictive performance, which reduces their clinical value in non-European populations and exacerbates healthcare disparities. This study, conducted by CGM Investigators Tian Ge, Hailiang Huang, Alicia Martin and colleagues, developed a computational framework, termed PRS-CSx, that can integrate genomic data from multiple populations to improve polygenic prediction in diverse populations. Leveraging large-scale global biobanks and disease-focused cohorts, the investigators showed that PRS-CSx substantially improved the prediction accuracy of biomarkers and disease risk in non-European populations. This work represents an important step towards the implementation of PRS into routine healthcare.

Childhood maltreatment increases the likelihood of mental health problems, but outcomes vary widely. The timing of maltreatment may help explain these outcome differences. Specifically, impacts might be stronger during sensitive periods in development when the brain is maximally sensitive to particular types of environmental input. Is there consensus about peak periods of sensitivity to childhood maltreatment? This review led by CGM investigator Erin Dunn, analysis of observational studies in humans found that most studies (75%, or 89 of 118 studies) reported timing effects, but did not identify consistent sensitive periods across studies. The investigators also examined study characteristics to inform a discussion of practices to support future research.

For the past 2 years, during the pandemic, CGM PI, Brian Skotko, and his multidisciplinary team set up and implemented a randomized control trial to see if they could design a multimodal behavioral intervention that could reduce COVID-19 infection, hospitalization, and death for those persons living in group homes in MA. They focused on people with serious mental illness (SMI) and people with intellectual and/or developmental disabilities (ID/DD)—and the staff who work with them in group homes. This “protocol paper” describes our unique design: a hybrid type 1 effective-implementation cluster randomized trial, which means that half of the 400 participating group homes got our tailored intervention (and the other half continued to receive the statewide recommendations). For the past 2 years, they have been implementing this trial, which has now come to a close.