Time trends in depression prevalence among Swedish 85-year-olds: repeated cross-sectional population-based studies in 1986, 2008, and 2015

Natural Language Processing and Social Determinants of Health in Mental Health Research: AI-Assisted Scoping Review

Background

The use of natural language processing (NLP) in mental health research is increasing, with a wide range of applications and datasets being investigated.

Objective

This review aims to summarize the use of NLP in mental health research, with a special focus on the types of text datasets and the use of social determinants of health (SDOH) in NLP projects related to mental health.

Methods

The search was conducted in September 2024 using a broad search strategy in PubMed, Scopus, and CINAHL Complete. All citations were uploaded to Covidence (Veritas Health Innovation) software. The screening and extraction process took place in Covidence with the help of a custom large language model (LLM) module developed by our team. This LLM module was calibrated and tuned to automate many aspects of the review process.

Results

The screening process, assisted by the custom LLM, led to the inclusion of 1768 studies in the final review. Most of the reviewed studies (n=665, 42.8%) used clinical data as their primary text dataset, followed by social media datasets (n=523, 33.7%). The United States contributed the highest number of studies (n=568, 36.6%), with depression (n=438, 28.2%) and suicide (n=240, 15.5%) being the most frequently investigated mental health issues. Traditional demographic variables, such as age (n=877, 56.5%) and gender (n=760, 49%), were commonly extracted, while SDOH factors were less frequently reported, with urban or rural status being the most used (n=19, 1.2%). Over half of the citations (n=826, 53.2%) did not provide clear information on dataset accessibility, although a sizable number of studies (n=304, 19.6%) made their datasets publicly available.

Conclusions

This scoping review underscores the significant role of clinical notes and social media in NLP-based mental health research. Despite the clear relevance of SDOH to mental health, their underutilization presents a gap in current research. This review can be a starting point for researchers looking for an overview of mental health projects using text data. Shared datasets could be used to place more emphasis on SDOH in future studies.

Prevalence of depressive disorders among the very old in the 21st century

BACKGROUND

The aim of this study is to explore the prevalence of depressive disorders in very old adults over time, in rural/urban environments, between men/women, as well as to explore other factors associated with depressive disorders.

METHODS

This study was conducted utilizing the GERDA-database data, which consists of four cohorts of 85, 90 and 95+ year olds living in Northern Sweden. Participants could reside independently or in residential care. Data collections took place between 2000 and 2017. Descriptive data and logistic regression models were utilized to explore data.

RESULTS

The prevalence of depressive disorders increased between 2000/02 and 2015/17 in all age groups, with the highest percentages observed in the 95+ age group, reaching 53.6 % in 2015/17. The prevalence varied from 20.3 % in those without dementia to 65.1 % in those with dementia. Sex or living in an urban/rural environment was not associated with an increased risk of depression in the fully adjusted models. Dementia and reduced capacity in activities of daily living were associated with depressive disorders among 85 and 90-year-olds, while living alone was associated with depressive disorders in the 95+ age group.

LIMITATIONS

Potentially limited generalizability, as this study took place in northern Sweden.

CONCLUSIONS

The prevalence of depressive disorders among very old adults increases with age and the prevalence also increases throughout cohorts and time. These alarming rates of depressive disorders among the very old require immediate measures and further investigation. Future studies are needed to explore and monitor trends and to plan and design tailored interventions.

Prevalence and risk factors for acquired long QT syndrome in the emergency department: a retrospective observational study

BACKGROUND

Long QT syndrome (LQTS) is a heterogeneous syndrome that may be congenital or, more frequently, acquired. The real-world prevalence of acquired LQTS (aLQTS) in the emergency department (ED) remains to be determined. The aim of this study was to determine prevalence of aLQTS and its impact on symptoms on ED admissions.

METHODS

Electrocardiograms (ECG) of 5,056 consecutively patients admitted in the ED of a tertiary hospital between January 28th and March 17th of 2020 were reviewed. All patients with aLQTS were included. Clinical data with a focus on QT prolonging drugs and clinical factors were recorded. Statistical comparison was made between the groups with and without corrected QT (QTc) interval greater than 500 ms (value that is considered severely increased).

RESULTS

A total of 383 ECGs with prolonged QTc were recognized, corresponding to a prevalence of aLQTS at admission of 7.82%. Patients with aLQTS were more commonly men (53.3%) with an age of (73.49±14.79) years old and QTc interval of (505.3±32.4) ms. Only 20.4% of these patients with aLQTS were symptomatic. No ventricular arrhythmias were recorded. Patients with QT interval greater than 500 ms were more frequently female (59.5%; P<0.001) and were more frequently on QT prolonging drugs (77.3%; P=0.025). Main contributing factor was intake of antibiotics (odds ratio [OR] 4.680) followed by female gender (OR 2.473) and intake of antipsychotics (OR 1.925).

CONCLUSION

aLQTS is particularly prevalent in the ED. Female patients on antibiotics and antipsychotics are at particularly high risk. Efforts must be made to avoid, detect and treat aLQTS as early as possible.

Passive and active suicidal ideation in a population-based sample of older adults: Associations with polygenic risk scores of relevance for suicidal behavior

INTRODUCTION

There are few studies investigating genetic factors related to suicidal ideation or behavior in older adult populations. Our aim was to test associations between passive and active suicidal ideation and polygenic risk scores (PRSs) for suicidality and other traits of relevance for suicidality in old age (i.e. depression, neuroticism, loneliness, Alzheimer's disease, cognitive performance, educational attainment, and several specified vascular diseases) in a population-based sample aged 70 years and older.

METHODS

Participants in the prospective H70 study in Gothenburg, Sweden, took part in a psychiatric examination that included the Paykel questions on active and passive suicidal ideation. Genotyping was performed with the Neurochip (Illumina). After quality control of the genetic data the sample included 3467 participants. PRSs for suicidality and other related traits were calculated based on summary statistics from recent GWASs of relevance. Exclusion of persons with dementia or incomplete data on suicidal ideation yielded 3019 participants, age range 70-101 years. Associations between past year suicidal ideation (any level) and selected PRSs were analysed using general estimation equation (GEE) models, adjusted for sex and age.

RESULTS

We observed associations between passive/active suicidal ideation and PRSs for depression (three versions), neuroticism, and general cognitive performance. After excluding individuals with current major depressive disorder (MDD), similar associations were seen with PRS for neuroticism, general cognitive performance and two PRSs for depression. No associations were found between suicidal ideation and PRSs for suicidality, loneliness, Alzheimer's disease, educational attainment, or vascular disease.

DISCUSSION

Our results could indicate which types of genetic susceptibility that are of importance for suicidality in old age, and these findings can help to shed light on potential mechanisms that may be involved in passive and active suicidal ideation in late-life, also in those with no current MDD. However, due to the limited sample size, the results need to be interpreted with caution until replicated in larger samples.