Davide Morelli

Commonly used methods for assessing cardiac interoceptive accuracy have been criticised for assuming that all individuals perceive their heartbeat at the same delay following contraction of the heart. However, it remains unclear whether some individuals perceive their heartbeat at a particular phase of their cardiac cycle – that is, at a relative point in the cycle that may vary in absolute timing depending on heart rate - rather than at a specific delay. In the current study, data from 531 participants who had completed the Phase Adjustment Task (PAT) as a measure of cardiac interoceptive accuracy were analysed. In this task, participants are asked to adjust a virtual dial until tones appear synchronous with their heartbeats. Data were analysed using a novel framework that allows differentiation between delay-based and phase-based response patterns. 17% of interoceptive individuals demonstrated response patterns consistent with phase-based responding. These novel findings challenge current assumptions regarding individual differences in the perception of heartbeats, and suggest that many commonly used measures may underestimate the true proportion of heartbeat perceivers.

Risk stratification is an essential tool in the fight against many diseases, including chronic kidney disease. Recent work has focused on applying techniques from machine learning and leveraging the information contained in a patient’s electronic health record (EHR). Irregular intervals between data entries and the large number of variables tracked in EHR datasets can make them challenging to work with. Many of the difficulties associated with these datasets can be overcome by using large language models, such as bidirectional encoder representations from transformers (BERT). Previous attempts to apply BERT to EHR for risk stratification have shown promise. In this work we propose HERBERT, a novel application of BERT to EHR data. We identify two key areas where BERT models must be modified to adapt them to EHR data, namely: the embedding layer and the pretraining task. We show how changes to these can lead to improved performance, relative to the previous state of the art. We evaluate our model by predicting the transition of chronic kidney disease patients to end stage renal disease. The strong performance of our model justifies our architectural changes and suggests that large language models could play an important role in future renal risk stratification.

Background: Previous research suggests that the processing of internal body sensations (interoception) affects how we experience pain. There is some evidence that people with fibromyalgia syndrome (FMS), which is a condition characterised by chronic pain and fatigue, may have altered interoceptive processing. However, extant findings are inconclusive, and some tasks previously used to measure interoception are of questionable validity. We used a task which overcomes problems with previous tasks – the Phase Adjustment Task (PAT) – to measure interoception in adults with FMS. Methods: We examined: (i) the tolerability of the PAT in an FMS sample (N = 154); (ii) if there are differences in facets of interoception (PAT performance, PAT-related confidence, and scores on the Private Body Consciousness Scale) between an FMS sample and an age- and gender-matched pain-free control sample (N = 94); and (iii) if subgroups of participants with FMS could be identified according to interoceptive accuracy levels. Results: After including additional task breaks and a recommended hand posture, we found that the PAT was tolerable in the FMS sample. Participants in the FMS sample were more likely to be classified as ‘interoceptive’ on the PAT, and had significantly higher self-reported interoception and interoceptive beliefs compared to participants in the pain-free sample. Within the FMS sample, participants who were classified as interoceptive on the PAT had significantly lower symptom impact than the unclassified participants. Conversely, self-reported interoception was positively correlated with FMS symptom severity and impact. Conclusions: The present findings suggest that interoception may be an important factor to consider in understanding and managing FMS symptoms, and that the PAT is a useful tool for assessing interoception in this population. We recommend future longitudinal work to better understand associations between fluctuating FMS symptoms and interoceptive processing.

BACKGROUND Resilience is thought to be associated with how individuals perceive, experience, and react to stressful situations. Most previous research has focused on the psychosocial and behavioral aspects of resilience. More recently, researchers have started to investigate potential biological markers of resilience e.g., heart rate variability (HRV) and blood pressure (BP). OBJECTIVE The main objective of this research study was to investigate whether resilience influences cardiovascular recovery following exposure to acute stress. METHODS Physiological markers of stress (BP, heart rate (HR) and HRV) were collected to evaluate whether resilience affects stress recovery. Participants (N=72) completed a series of questionnaires (resilience, stress, psychological distress, stressful life events, anxiety, and affect) and two mental stress tasks: a mock interview task and an arithmetic task. A recovery period followed immediately after. HRV and HR were assessed continuously throughout the research study and participants BP was assessed at three separate instances. The physiological markers were divided into three sections of at least 5 minutes of recordings (HRV and HR) for this study including: baseline (before mental stress tasks), stress period (during the mental stress tasks) and the recovery period (after mental stress tasks). RESULTS The experimental manipulation was successful as participants rated their feelings of pressure higher after stress exposure (M=4.2) than before (M=3.0, P=.04). There were also significant increases in HR between baseline (M=79.9) and the last three minutes of the mental stress tasks (M=84.5, P< .001). Following stress exposure, the high resilience group demonstrated enhanced BP and HRV recovery (Root Mean Square of the Successive Differences, RMSSD) relative to the low resilience group (P<.001, P=.04 respectively). The model was not significant for HR or high frequency (HF), an alternative measure of HRV which indicates parasympathetic activity (P=.82, P=.22 respectively). CONCLUSIONS The findings suggested that the effects of resilience are only exhibited during recovery, following a stress episode. Those with high resilience showed enhanced BP and RMSSD recovery after acute stress. The results suggested that HRV (RMSSD) could be an indicator of resilience and a protective factor for both mental and physical health. This has implications for both interventions, future research and shows how important resilience is in physiological stress recovery. CLINICALTRIAL The research was pre-registered on the Open Science Framework (osf.io/rh8dw).

Growing evidence suggests that poor sleep health is associated with cardiovascular risk. However, research in this area often relies upon recollection dependent questionnaires or diaries. Accelerometers provide an alternative tool for measuring sleep parameters objectively. This study examines the association between wrist-worn accelerometer-derived sleep onset timing and cardiovascular disease (CVD). We derived sleep onset and waking up time from accelerometer data collected from 103 712 UK Biobank participants over a period of 7 days. From this, we examined the association between sleep onset timing and CVD incidence using a series of Cox proportional hazards models. A total of 3172 cases of CVD were reported during a mean follow-up period of 5.7 (±0.49) years. An age- and sex-controlled base analysis found that sleep onset time of 10:00 p.m.–10:59 p.m. was associated with the lowest CVD incidence. An additional model, controlling for sleep duration, sleep irregularity, and established CVD risk factors, did not attenuate this association, producing hazard ratios of 1.24 (95% confidence interval, 1.10–1.39; P < 0.005), 1.12 (1.01–1.25; P = 0.04), and 1.25 (1.02–1.52; P = 0.03) for sleep onset <10:00 p.m., 11:00 p.m.–11:59 p.m., and ≥12:00 a.m., respectively, compared to 10:00 p.m.–10:59 p.m. Importantly, sensitivity analyses revealed this association with increased CVD risk was stronger in females, with only sleep onset <10:00 p.m. significant for males. Our findings suggest the possibility of a relationship between sleep onset timing and risk of developing CVD, particularly for women. We also demonstrate the potential utility of collecting information about sleep parameters via accelerometry-capable wearable devices, which may serve as novel cardiovascular risk indicators.

Background All-cause mortality (ACM) scores are a useful tool for identifying individuals with decreased life expectancy. An interpretable score consisting of smartphone-obtainable variables could allow for long-term management of individual health and support the next generation of healthcare monitoring and preventative practices. The aim of this study was to develop a 10-year ACM risk score using the UK Biobank dataset, using only digitally-obtainable variables. Methods The models were developed using the full UK Biobank cohort comprising nearly 500,000 individuals. We extracted 399 features from the dataset and, through a data-driven feature selection process with subsequent clinical review, identified 34 features for the final model. As part of the study, we compared two survival analysis approaches: Cox proportional hazards model and DeepSurv, a deep learning-based survival analysis algorithm. Results Before feature selection, Cox performed similarly to DeepSurv, achieving a c-index of 0.771 (95% CI 0.770–0.772) and 0.774 (95% CI 0.772–0.775) on the test dataset, respectively. Using the selected 34 features, the c-index of Cox decreased slightly to 0.770 (95% CI 0.769–0.770) and DeepSurv to 0.758 (95% CI 0.755–0.762). The models show excellent calibration at 10 years. Conclusions This study improves on a previous smartphone-compatible score, C-Score, by incorporating non-modifiable factors in addition to variables which can be actively modified to reduce risk. This score is comprehensive, easily interpretable and actionable, and as such, could provide a powerful tool for preventative healthcare.

Aims: Growing evidence suggests that sleep quality is associated with cardiovascular risk. However, research in this area often relies upon recollection dependant questionnaires or diaries. Accelerometers provide an alternative tool for deriving sleep parameters measuring sleep patterns objectively. This study examines the associations between accelerometer derived sleep onset timing and cardiovascular disease (CVD). Methods and Results: We derived sleep onset and waking up time from accelerometer data collected from 103,712 UK Biobank participants over a period of seven days. From this, we examined the association between sleep onset timing and CVD incidence using a series of Cox proportional hazards models. 3172 cases of CVD were reported during a mean follow-up period of 5·7 (±0·49) years. An age- and sex-controlled base analysis found that sleep onset time of 10:00pm-10:59pm was associated with the lowest CVD incidence. A fully adjusted model, additionally controlling for sleep duration, sleep irregularity, and established CVD risk factors, was unable to eliminate this association, producing hazard ratios of 1·24 (95% CI, 1·10–1·39; p p= 0·04), and 1·25 (1·02–1·52; p= 0·03) for sleep onset <10:00pm, 11:00pm-11:59pm, and ≥12:00am, respectively, compared to 10:00pm-10:59pm. Importantly, sensitivity analyses revealed this association was stronger in females, with only sleep onset <10:00pm significant for males. Conclusions: Our findings suggest an independent relationship between sleep onset timing and risk of developing CVD, particularly for women. We also demonstrate the potential utility of collecting information about sleep parameters via accelerometry-capable wearable devices, which may serve as novel cardiovascular risk indicators. Translational Perspective: Abnormal sleep timing is a feature of poor-quality sleep and is likely to be associated with cardiovascular diseases. Accelerometer-derived measures of sleep quality, particularly sleep timing, may be associated with greater vulnerability to cardiovascular disease, particularly in women. Onset of sleep timing may constitute a novel cardiovascular risk factor and target for preventative intervention. Wearable devices equipped with accelerometers may be useful in assessing sleep related cardiovascular disease risk factors as they would allow passive measurement of sleep parameters, such as sleep timing.

Aims: Cardiovascular diseases (CVDs) are among the leading causes of death worldwide. Predictive scores providing personalized risk of developing CVD are increasingly used in clinical practice. Most scores, however, utilize a homogenous set of features and require the presence of a physician. The aim was to develop a new risk model (DiCAVA) using statistical and machine learning techniques that could be applied in a remote setting. A secondary goal was to identify new patient-centric variables that could be incorporated into CVD risk assessments. Methods and results : Across 466 052 participants, Cox proportional hazards (CPH) and DeepSurv models were trained using 608 variables derived from the UK Biobank to investigate the 10-year risk of developing a CVD. Data-driven feature selection reduced the number of features to 47, after which reduced models were trained. Both models were compared to the Framingham score. The reduced CPH model achieved a c-index of 0.7443, whereas DeepSurv achieved a c-index of 0.7446. Both CPH and DeepSurv were superior in determining the CVD risk compared to Framingham score. Minimal difference was observed when cholesterol and blood pressure were excluded from the models (CPH: 0.741, DeepSurv: 0.739). The models show very good calibration and discrimination on the test data. Conclusion: We developed a cardiovascular risk model that has very good predictive capacity and encompasses new variables. The score could be incorporated into clinical practice and utilized in a remote setting, without the need of including cholesterol. Future studies will focus on external validation across heterogeneous samples.

UNSTRUCTURED Interoception has increasingly been the focus of psychiatric research, due to its hypothesised role in mental health. Existing interoceptive tasks are often burdensome and require specialised equipment, limiting their usage in vulnerable populations. We report on the development of the CARdiac Elevation Detection (CARED) task. Participants’ heart rate is recorded by a wearable device connected to a mobile application. Notifications are sent to participants’ mobile throughout the day over a period of 4 weeks. Participants are asked to state whether their heart rate is higher than usual, rate their confidence and describe the activity they were involved in when the notification occurred. Data (N=30) revealed that ⅓ of the sample was classified as interoceptive and that participants presented overall good insight into their interoceptive abilities. Given its ease of administration and accessibility, the CARED task has the potential to be a significant asset for psychiatric and developmental research.