Development and validation of a machine learning model for prediction of type 2 diabetes in patients with mental illness.

BACKGROUND: Type 2 diabetes (T2D) is approximately twice as common among individuals with mental illness compared with the background population, but may be prevented by early intervention on lifestyle, diet, or pharmacologically. Such prevention relies on identification of those at elevated risk (prediction). The aim of this study was to develop and validate a machine learning model for prediction of T2D among patients with mental illness. METHODS: The study was based on routine clinical data from electronic health records from the psychiatric services of the Central Denmark Region. A total of 74,880 patients with 1.59 million psychiatric service contacts were included in the analyses. We created 1343 potential predictors from 51 source variables, covering patient-level information on demographics, diagnoses, pharmacological treatment, and laboratory results. T2D was operationalised as HbA1c ≥48 mmol/mol, fasting plasma glucose ≥7.0 mmol/mol, oral glucose tolerance test ≥11.1 mmol/mol or random plasma glucose ≥11.1 mmol/mol. Two machine learning models (XGBoost and regularised logistic regression) were trained to predict T2D based on 85% of the included contacts. The predictive performance of the best performing model was tested on the remaining 15% of the contacts. RESULTS: The XGBoost model detected patients at high risk 2.7 years before T2D, achieving an area under the receiver operating characteristic curve of 0.84. Of the 996 patients developing T2D in the test set, the model issued at least one positive prediction for 305 (31%). CONCLUSION: A machine learning model can accurately predict development of T2D among patients with mental illness based on routine clinical data from electronic health records. A decision support system based on such a model may inform measures to prevent development of T2D in this high-risk population.

Lexical stability of psychiatric clinical notes from electronic health records over a decade.

OBJECTIVE: Natural language processing (NLP) methods hold promise for improving clinical prediction by utilising information otherwise hidden in the clinical notes of electronic health records. However, clinical practice - as well as the systems and databases in which clinical notes are recorded and stored - change over time. As a consequence, the content of clinical notes may also change over time, which could degrade the performance of prediction models. Despite its importance, the stability of clinical notes over time has rarely been tested. METHODS: The lexical stability of clinical notes from the Psychiatric Services of the Central Denmark Region in the period from January 1, 2011, to November 22, 2021 (a total of 14,811,551 clinical notes describing 129,570 patients) was assessed by quantifying sentence length, readability, syntactic complexity and clinical content. Changepoint detection models were used to estimate potential changes in these metrics. RESULTS: We find lexical stability of the clinical notes over time, with minor deviations during the COVID-19 pandemic. Out of 2988 data points, 17 possible changepoints (corresponding to 0.6%) were detected. The majority of these were related to the discontinuation of a specific note type. CONCLUSION: We find lexical and syntactic stability of clinical notes from psychiatric services over time, which bodes well for the use of NLP for predictive modelling in clinical psychiatry.

Stability of diagnostic coding of psychiatric outpatient visits across the transition from the second to the third version of the Danish National Patient Registry.

OBJECTIVE: In Denmark, data on hospital contacts are reported to the Danish National Patient Registry (DNPR). The ICD-10 main diagnoses from the DNPR are often used as proxies for mental disorders in psychiatric research. With the transition from the second version of the DNPR (DNPR2) to the third (DNPR3) in February-March 2019, the way main diagnoses are coded in relation to outpatient treatment changed substantially. Specifically, in the DNPR2, each outpatient treatment course was labelled with only one main diagnosis. In the DNPR3, however, each visit during an outpatient treatment course is labelled with a main diagnosis. We assessed whether this change led to a break in the diagnostic time-series represented by the DNPR, which would pose a threat to the research relying on this source. METHODS: All main diagnoses from outpatients attending the Psychiatric Services of the Central Denmark Region from 2013 to 2021 (n = 100,501 unique patients) were included in the analyses. The stability of the DNPR diagnostic time-series at the ICD-10 subchapter level was examined by comparing means across the transition from the DNPR2 to the DNPR3. RESULTS: While the proportion of psychiatric outpatients with diagnoses from some ICD-10 subchapters changed statistically significantly from the DNPR2 to the DNPR3, the changes were small in absolute terms (e.g., +0.6% for F2-psychotic disorders and +0.6% for F3-mood disorders). CONCLUSION: The change from the DNPR2 to the DNPR3 is unlikely to pose a substantial threat to the validity of most psychiatric research at the diagnostic subchapter level.

Men with high serotonin 1B receptor binding respond to provocations with heightened amygdala reactivity.

Serotonin signalling influences amygdala reactivity to threat-related emotional facial expressions in healthy adults, but in vivo serotonin signalling has never been investigated in the context of provocative stimuli in aggressive individuals. The aim of this study was to evaluate associations between serotonin 1B receptor (5-HT1BR) levels and brain reactivity to provocations. We quantified regional 5-HT1BR binding using [11C]AZ10419369 positron emission tomography (PET) and measured brain activation following provocations with functional magnetic resonance imaging (fMRI) in eighteen violent offenders and 25 healthy control subjects. The point-subtraction aggression paradigm (PSAP) was used in fMRI to elicit provocations in terms of monetary subtractions from a fictive opponent. We estimated global 5-HT1BR binding using a linear structural equation model, with a single latent response variable (LV1B) modelling shared correlation between 5-HT1BR binding across multiple brain regions (neocortex, anterior and posterior cingulate cortex, raphe, amygdala, hippocampus and striatum). We tested whether the LV1B was associated with amygdala, striatal and prefrontal reactivity to provocations, adjusting for age, injected mass and group. Across participants, LV1B was statistically significantly positively associated with amygdala (p = 0.01) but not with striatal (p = 0.2) or prefrontal reactivity to provocations (p = 0.3). These findings provide novel evidence that 5-HT1BR levels are linked to amygdala reactivity to provocations in a cohort of men displaying a wide range of aggressive behavior. The data suggest that 5-HT1BR represents an intriguing target for reducing excessive neural reactivity to provocations and thereby putatively violent behavior.

Serotonin 1B Receptor Binding Is Associated With Trait Anger and Level of Psychopathy in Violent Offenders.

BACKGROUND: The involvement of serotonin in aggression has traditionally been attributed to impaired prefrontal serotonergic inhibitory control of emotional reactions to provocations in antisocial individuals. However, it is unclear which specific serotonergic receptors are involved in the effects. A large body of preclinical research supports a specific role of serotonin 1B receptors (5-HT1BRs) in aggression and impulsivity, but this has never been evaluated in humans. METHODS: Nineteen incarcerated violent offenders and 24 healthy control nonoffenders were included and examined with positron emission tomography, using the radioligand [11C]AZ10419369 for quantification of cerebral 5-HT1BR binding in three regions of interest: the anterior cingulate cortex, orbitofrontal cortex, and striatum. RESULTS: Group status significantly moderated the association between striatal 5-HT1BRs and trait anger (difference in slopes, pcorrected = .04). In the violent offender group, striatal 5-HT1BR binding was positively correlated with self-reported trait anger (p = .0004), trait psychopathy (p = .008), and level of psychopathy according to the Psychopathy Checklist-Revised (p = .02). We found no group differences in 5-HT1BR binding. CONCLUSIONS: Our data demonstrate for the first time in humans a specific involvement of 5-HT1BR binding in anger and psychopathy. 5-HT1BRs putatively represent a molecular target for development of pharmacologic antiaggressive treatments.

Testosterone levels in healthy men correlate negatively with serotonin 4 receptor binding.

The serotonergic system integrates sex steroid information and plays a central role in mood and stress regulation, cognition, appetite and sleep. This interplay may be critical for likelihood of developing depressive episodes, at least in a subgroup of sensitive individuals. The serotonin 4 receptor (5-HT4R) indexes central serotonergic tonus, which may be related to endogenous sex-steroid levels in the mentally healthy state even though this remains elusive. Here we evaluate if peripheral levels of estradiol and testosterone are associated with 5-HT4R binding as imaged by [11C]SB207145 positron emission tomography in a group of 41 healthy men. We estimated global 5-HT4R binding using a latent variable model framework, which models shared correlation between 5-HT4R across multiple brain regions (hippocampus, amygdala, posterior and anterior cingulate, thalamus, pallidostriatum and neocortex). We tested whether testosterone and estradiol predict global 5-HT4R, adjusting for age. We found that testosterone, but not estradiol, correlated negatively with global 5-HT4R levels (p=0.02) suggesting that men with high levels of testosterone have higher cerebral serotonergic tonus. Our findings corroborate the link between sex hormone levels and serotonin signalling. Future longitudinal studies in clinical relevant populations are needed to elucidate the potential importance of testosterone in the pathophysiology of e.g. major depression and its treatment.

Violent offenders respond to provocations with high amygdala and striatal reactivity.

The ability to successfully suppress impulses and angry affect is fundamental to control aggressive reactions following provocations. The aim of this study was to examine neural responses to provocations and aggression using a laboratory model of reactive aggression. We used a novel functional magnetic resonance imaging point-subtraction aggression paradigm in 44 men, of whom 18 were incarcerated violent offenders and 26 were control non-offenders. We measured brain activation following provocations (monetary subtractions), while the subjects had the possibility to behave aggressively or pursue monetary rewards. The violent offenders behaved more aggressively than controls (aggression frequency 150 vs 84, P = 0.03) and showed significantly higher brain reactivity to provocations within the amygdala and striatum, as well as reduced amygdala-prefrontal and striato-prefrontal connectivity. Amygdala reactivity to provocations was positively correlated with task-related behavior in the violent offenders. Across groups, striatal and prefrontal reactivity to provocations was positively associated with trait anger and trait aggression. These results suggest that violent individuals display abnormally high neural sensitivity to social provocations, a sensitivity related to aggressive behavior. These findings provide novel insight into the neural pathways that are sensitive to provocations, which is critical to more effectively shaped interventions that aim to reduce pathological aggressive behavior.