Multidomain Social Determinants of Depressive Symptoms for the Elderly with Chronic Diseases: Evidence from the China Health and Retirement Longitudinal Study

Elderly individuals with chronic diseases (CDs) have a much higher risk of mental disorders, especially depression. This study aimed to identify the multidomain social determinants of occurrence and degree of depressive symptoms for the Chinese elderly with CDs. Data of 3438 elderly individuals (aged over 60 years) with CDs were drawn from the fourth wave of the China Health and Retirement Longitudinal Study implemented in 2018. Logistic regression was used to describe associations with the occurrence of depressive symptoms within and across multidomain social determinants (demographic, economic, neighborhood, environmental, and social and cultural). The Shapley value decomposition method was used to measure the relative importance of variables of the five domains. A quantile regression model was used to test how the effects of social factors vary across different points of depression score distributions. Approximately 40.1% of Chinese elderly individuals with CDs reported depressive symptoms. Respondents who were female, had a low income, experienced a disability, lived in rural areas, and were not engaged in work had a higher probability of suffering from depressive symptoms. Conversely, increased age, being covered by social security and being well-educated had a protective effect. Data also showed that the effects of these associated factors varied across different points of depression score distributions. The fact that socially disadvantaged people are more vulnerable to severe depressive symptoms implies that psychological health services and intervention strategies should target this population.

Pre-Statistical Considerations for Harmonization of Cognitive Instruments: Harmonization of ARIC, CARDIA, CHS, FHS, MESA, and NOMAS

Background: Meta-analyses of individuals’ cognitive data are increasing to investigate the biomedical, lifestyle, and sociocultural factors that influence cognitive decline and dementia risk. Pre-statistical harmonization of cognitive instruments is a critical methodological step for accurate cognitive data harmonization, yet specific approaches for this process are unclear. Objective: To describe pre-statistical harmonization of cognitive instruments for an individual-level meta-analysis in the blood pressure and cognition (BP COG) study. Methods: We identified cognitive instruments from six cohorts (the Atherosclerosis Risk in Communities Study, Cardiovascular Health Study, Coronary Artery Risk Development in Young Adults study, Framingham Offspring Study, Multi-Ethnic Study of Atherosclerosis, and Northern Manhattan Study) and conducted an extensive review of each item’s administration and scoring procedures, and score distributions. Results: We included 153 cognitive instrument items from 34 instruments across the six cohorts. Of these items, 42%were common across ≥2 cohorts. 86%of common items showed differences across cohorts. We found administration, scoring, and coding differences for seemingly equivalent items. These differences corresponded to variability across cohorts in score distributions and ranges. We performed data augmentation to adjust for differences. Conclusion: Cross-cohort administration, scoring, and procedural differences for cognitive instruments are frequent and need to be assessed to address potential impact on meta-analyses and cognitive data interpretation. Detecting and accounting for these differences is critical for accurate attributions of cognitive health across cohort studies.

Preoperative Psychometric Properties of PROMIS Upper Extremity, Pain Interference, and Depression in Bankart Repair and Rotator Cuff Repair (115)

Objectives: To examine the preoperative and postoperative performance of PROMIS Upper Extremity 2.0 across various orthopedic procedures for the upper extremity. Secondarily, to define susceptibility to pre- and post-operative floor and ceiling effects. Methods: Retrospective analysis of prospectively collected patient-reported outcome (PRO) data was conducted utilizing an electronic outcome registry for procedures between May 2017 and August 2018. Current procedural terminology (CPT) codes were utilized to examine cohorts for various upper extremity orthopedic procedures including Bankart repair and arthroscopic rotator cuff repair (ARCR). Shapiro-Wilks normality testing was used to assess score distributions for normalcy; given non-normal score distributions, Spearman correlation coefficients were calculated for preoperative patient-reported outcome (PRO) scores. Absolute floor and ceiling effects were calculated for each time point based on CPT code. Results: A total of 488 patients were included across Bankart repair (n=109, 29.3 + 12.5 years) and ARCR (n=379, 57.5 + 9.5 years) cohorts. In the Bankart repair cohort, PROMIS PI demonstrate strong correlations with ASES (r=-0.63), Constant (r=-0.75), PROMIS UE (r=-0.75), and the VR6D (r=-0.61). PROMIS Depression (r=0.23 vs. 0.17), SF12 MCS (r=0.34 vs. 0.11), and VR12 MCS (r=0.44 vs. 0.15) exhibited poor correlations with PROMIS PI and UE, respectively. In the ARCR cohort, PROMIS PI scores demonstrated a good correlation with PROMIS UE (r=0.61). Constant (r=0.58 vs. 0.67), VR12 PCS (r=0.58 vs. 0.47), and VR6D (r=0.67 vs. 0.53) exhibited good correlations with both PROMIS PI and UE measures, respectively. No significant absolute floor or ceiling effects were observed for the PROMIS instruments except PROMIS Depression; an absolute floor was noted for both Bankart (n=12, 30%) and ARCR (n=38, 14.7%) groups. Conclusions: PROMIS PI and UE instruments perform comparably to legacy outcome instruments in patients receiving Bankart repair, as well as those receiving ARCR. Furthermore, in both populations, the PROMIS Depression instrument exhibits absolute floor effects, whereas PROMIS PI and UE instruments fail to demonstrate any absolute floor or ceiling effects.

Video Summarization via Label Distributions Dual-Reward

Reinforcement learning maps from perceived state representation to actions, which is adopted to solve the video summarization problem. The reward is crucial for deal with the video summarization task via reinforcement learning, since the reward signal defines the goal of video summarization. However, existing reward mechanism in reinforcement learning cannot handle the ambiguity which appears frequently in video summarization, i.e., the diverse consciousness by different people on the same video. To solve this problem, in this paper label distributions are mapped from the CNN and LSTM-based state representation to capture the subjectiveness of video summaries. The dual-reward is designed by measuring the similarity between user score distributions and the generated label distributions. Not only the average score but also the the variance of the subjective opinions are considered in summary generation. Experimental results on several benchmark datasets show that our proposed method outperforms other approaches under various settings.

Assessing quality of life in pediatric fibrous dysplasia and McCune Albright syndrome: PEDS-QL and HADS data from the Fibrous Dysplasia Foundation Patient Registry

Purpose The International FD/MAS Consortium recently encouraged using the Pediatric Quality of Life Inventory (PEDS-QL) and the Hospital Anxiety and Depression scales (HADS) in clinical care. This study examines scores on these measures among pediatric fibrous dysplasia and McCune Albright (FD/MAS) patients to initiate consideration of their use in clinical treatment. Methods This is a retrospective analysis of pediatric data from 39 minors, ages 2–17, entered in the Fibrous Dysplasia Foundation Patient Registry from July 2016 to December 2018. Sample means and score distributions are compared to general population and chronic disease benchmarks. Associations with medical and demographic variables are also explored. Results Mean PEDS-QL scores for children 2–7 were inconclusive in determining at risk status for impaired quality of life (QOL). Individual score distributions suggested up to half experienced extensive physical or social impairment. Means and individual score distributions for the physical and psychosocial components of the PEDS-QL for children 8–17 suggested many were at risk of impaired QOL. Over half of 13–17 year-olds met the clinical benchmark for anxiety. Older males scored better than females on the PEDS-QL and HADS. Pain frequency was associated with physical function for older children. Conclusions Older children with FD/MAS may be more compromised in terms of psychosocial QOL than previously reported. Clinicians should be attentive to the influence of gender on QOL in older children. Online patient registries associated with rare diseases have the potential to serve as efficient and cost-effective mechanisms to jumpstart examination of new measures in consideration for clinical use.

New mixture models for decoy-free false discovery rate estimation in mass spectrometry proteomics

Motivation Accurate estimation of false discovery rate (FDR) of spectral identification is a central problem in mass spectrometry-based proteomics. Over the past two decades, target-decoy approaches (TDAs) and decoy-free approaches (DFAs) have been widely used to estimate FDR. TDAs use a database of decoy species to faithfully model score distributions of incorrect peptide-spectrum matches (PSMs). DFAs, on the other hand, fit two-component mixture models to learn the parameters of correct and incorrect PSM score distributions. While conceptually straightforward, both approaches lead to problems in practice, particularly in experiments that push instrumentation to the limit and generate low fragmentation-efficiency and low signal-to-noise-ratio spectra. Results We introduce a new decoy-free framework for FDR estimation that generalizes present DFAs while exploiting more search data in a manner similar to TDAs. Our approach relies on multi-component mixtures, in which score distributions corresponding to the correct PSMs, best incorrect PSMs and second-best incorrect PSMs are modeled by the skew normal family. We derive EM algorithms to estimate parameters of these distributions from the scores of best and second-best PSMs associated with each experimental spectrum. We evaluate our models on multiple proteomics datasets and a HeLa cell digest case study consisting of more than a million spectra in total. We provide evidence of improved performance over existing DFAs and improved stability and speed over TDAs without any performance degradation. We propose that the new strategy has the potential to extend beyond peptide identification and reduce the need for TDA on all analytical platforms. Availabilityand implementation https://github.com/shawn-peng/FDR-estimation. Supplementary information Supplementary data are available at Bioinformatics online.