Lessons learned from conducting mental health intervention research in schools in the global south: Our experiences in South Africa and Kenya

Most of the world's population of young people live in lower-and middle-income countries (LMICs; (Weine, Horvath Marques, Singh, & Pringle, 2020)), and these young people experience heightened rates of known risk factors for developing mental disorders such as poverty and exposure to trauma (Atwoli, Stein, Koenen, & McLaughlin, 2015). Access to professional psychological treatments is limited in LMICs due to structural barriers (e.g., a dearth of trained professionals) and cultural factors like stigma and beliefs about mental health and illness. Therefore, schools, which are widely attended, may be a good location for providing mental health interventions, and it is important that we develop and evaluate feasible, acceptable, effective, and scalable interventions for use in this context. Yet under 10% of clinical trials of psychotherapies (Venturo-Conerly, Eisenman, Wasil, Singla, & Weisz, 2022) have been conducted in LMICs. And there are particular challenges to conducting research in schools, as has been highlighted in the UK context by Moore et al. (2022). Building on that commentary, our aim herein is to share our learnings from conducting psychotherapy research in schools in Kenya and South Africa.

Interventions to address loneliness and social isolation in young people: A systematic review of the evidence on acceptability and effectiveness

INTRODUCTION

Loneliness is prevalent and associated with negative health outcomes in young people. Our understanding of how it can be best addressed is limited. This systematic review aims to assess the acceptability and effectiveness of interventions to reduce and prevent loneliness and social isolation in young people.

METHODS

Six bibliographic databases were searched; references of included studies were screened for relevant literature. A pre-defined framework was used for data extraction. Quality appraisal was performed using the Mixed Method Appraisal Tool. Data were synthesised narratively.

RESULTS

9,358 unique references were identified; 28 publications from 16 interventions met the inclusion criteria. The majority of interventions were high intensity, individual or small group interventions, often targeted at specific 'at risk' populations. While 14 interventions were associated with a statistically significant reduction in loneliness or social isolation, the heterogeneous measures of loneliness, small sample sizes, short periods of follow-up and high attrition rates limit evidence on effectiveness. Interventions implemented in more general populations of young people appeared more acceptable than those in specific 'at risk' populations.

CONCLUSION

High intensity interventions are unlikely to be feasible at a population level. Further work is required to develop and evaluate theoretically-informed loneliness interventions for young people that reach wider audiences.

The sustainability of public health interventions in schools: a systematic review

BACKGROUND

The sustainability of school-based health interventions after external funds and/or other resources end has been relatively unexplored in comparison to health care. If effective interventions discontinue, new practices cannot reach wider student populations and investment in implementation is wasted. This review asked: What evidence exists about the sustainability of school-based public health interventions? Do schools sustain public health interventions once start-up funds end? What are the barriers and facilitators affecting the sustainability of public health interventions in schools in high-income countries?

METHODS

Seven bibliographic databases and 15 websites were searched. References and citations of included studies were searched, and experts and authors were contacted to identify relevant studies. We included reports published from 1996 onwards. References were screened on title/abstract, and those included were screened on full report. We conducted data extraction and appraisal using an existing tool. Extracted data were qualitatively synthesised for common themes, using May's General Theory of Implementation (2013) as a conceptual framework.

RESULTS

Of the 9677 unique references identified through database searching and other search strategies, 24 studies of 18 interventions were included in the review. No interventions were sustained in their entirety; all had some components that were sustained by some schools or staff, bar one that was completely discontinued. No discernible relationship was found between evidence of effectiveness and sustainability. Key facilitators included commitment/support from senior leaders, staff observing a positive impact on students' engagement and wellbeing, and staff confidence in delivering health promotion and belief in its value. Important contextual barriers emerged: the norm of prioritising educational outcomes under time and resource constraints, insufficient funding/resources, staff turnover and a lack of ongoing training. Adaptation of the intervention to existing routines and changing contexts appeared to be part of the sustainability process.

CONCLUSIONS

Existing evidence suggests that sustainability depends upon schools developing and retaining senior leaders and staff that are knowledgeable, skilled and motivated to continue delivering health promotion through ever-changing circumstances. Evidence of effectiveness did not appear to be an influential factor. However, methodologically stronger primary research, informed by theory, is needed.

TRIAL REGISTRATION

The review was registered on PROSPERO: CRD42017076320, Sep. 2017.

ID: 135: SINGLE CELL GENE EXPRESSION STUDIES IN LUPUS MONOCYTES REVEAL A UNIQUE ANTII-INFLAMMATORY NON-CLASSICAL MONOCYTE POPULATION ASSOCIATED WITH CLINICAL QUIESCENCE

Background

Our previous studies have shown that different cell types from the same blood sample demonstrate diverse gene expression parameters. In follow up work, it seems that this diversity extends to cells of the same type from the same blood sample. In this study, we examine single cell gene expression in SLE patient monocytes and determine correlations with clinical features.

Methods

CD14++CD16− classical monocytes (CLs) and CD14dimCD16+ non-classical monocytes (NCLs) from SLE patients were purified by magnetic separation. The Fluidigm single cell capture and pre-amplification system was used for single cell capture and target gene pre-amplification. Fluidigm Biomark system (Rt-PCR system) was used to quantify expression of 87 monocyte-related genes. IFN-induced genes in monocytes were identified by culturing monocytes isolated from whole blood of healthy controls with or without IFN-α. Genes significant up-regulated by IFN were identified as IFN-induced genes in current study. An individual cell IFN score was given based upon the sum of expression of IFN-induced genes.

Results

Both CLs and NCLs demonstrated a wide range of expression of IFN-induced genes, and NCL monocytes had higher IFN scores than CL monocytes. Using unsupervised hierarchical clustering, we found four gene sets that clustered monocytes functionally. These included an IFN-induced gene set, two inflammatory gene sets, and one immunosuppressive gene set. Interestingly, we could define a large subset of NCL monocytes with upregulation of suppressive transcripts (including TGF-β and PDL1) and IFN-induced transcripts were also upregulated, while the two inflammatory gene sets were down-regulated. These cells were highly over-represented in a patient with inactive disease who was on immunosuppressants at the time of blood draw. The proportion of anti-inflammatory gene set expressing NCLs was inversely correlated with anti-dsDNA titers (rho=−0.77, p=0.0051) and positively correlated with C3 complement (rho=0.68, p=0.030) in the SLE patient group, suggesting that these cells are also associated with serological quiescence.

Conclusion

Using single cell gene expression, we have identified a unique population of NCL monocytes in SLE patients with upregulation of a combination of anti-inflammatory and IFN-induced transcripts. These cells correspond with clinical and serological quiescence.

Fibroblast Biomarkers of Sporadic Parkinson's Disease and LRRK2 Kinase Inhibition

It has been uncertain whether specific disease-relevant biomarker phenotypes can be found using sporadic Parkinson’s disease (PD) patient-derived samples, as it has been proposed that there may be a plethora of underlying causes and pathological mechanisms. Fibroblasts derived from familial PD patients harboring leucine-rich repeat kinase 2 (LRRK2), PTEN-induced putative kinase 1 (PINK1), and Parkin mutations show clear disease-relevant mitochondrial phenotypes, which are exacerbated under conditions of pharmacological stress. We utilized fibroblasts derived from non-familial sporadic PD patients (without LRRK2 mutations) or LRRK2 mutation carriers to directly compare the cellular phenotypes during and after mitochondrial stress. We then determined the effects of pharmacological LRRK2 kinase inhibition using LRRK2-in-1. We found that there were two distinct populations of sporadic PD patient-derived fibroblast lines. One group of sporadic PD lines was highly susceptible to valinomycin-induced mitochondrial depolarization, emulating the mutant LRRK2 phenotype. These lines showed elevated mitochondrial superoxide/ nitric oxide levels, displayed increased mitochondrial and lysosome co-localization, and an increased rate of mitochondrial collapse, which corresponded with changes in mitochondrial fission and fusion proteins. The application of LRRK2-in-1 reversed decreased levels of mitochondrial and lysosome co-localization and partially restored mitochondrial network associated proteins and the mitochondrial membrane potential in the fibroblasts. This study identifies novel mitochondrial biomarkers in sporadic PD patient-derived fibroblast lines, which could be used as preclinical tools in which to test novel and known neuroprotective compounds.

Comparison of a Brassica oleracea genetic map with the genome of Arabidopsis thaliana

Brassica oleracea is closely related to the model plant, Arabidopsis thaliana. Despite this relationship, it has been difficult to both identify the most closely related segments between the genomes and determine the degree of genome replication within B. oleracea relative to A. thaliana. These difficulties have arisen in part because both species have replicated genomes, and the criteria used to identify orthologous regions between the genomes are often ambiguous. In this report, we compare the positions of sequenced Brassica loci with a known position on a B. oleracea genetic map to the positions of their putative orthologs within the A. thaliana genome. We use explicit criteria to distinguish orthologous from paralogous loci. In addition, we develop a conservative algorithm to identify collinear loci between the genomes and a permutation test to evaluate the significance of these regions. The algorithm identified 34 significant A. thaliana regions that are collinear with >28% of the B. oleracea genetic map. These regions have a mean of 3.3 markers spanning 2.1 Mbp of the A. thaliana genome and 2.5 cM of the B. oleracea genetic map. Our findings are consistent with the hypothesis that the B. oleracea genome has been highly rearranged since divergence from A. thaliana, likely as a result of polyploidization.