BACKGROUND Dementia has become a major public health concern due to its heavy disease burden. Mild cognitive impairment (MCI) is a transitional stage between healthy aging and dementia. Early identification of MCI is an essential step in dementia prevention. OBJECTIVE Based on machine learning (ML) methods, this study aimed to develop and validate a stable and scalable panel of cognitive tests for the early detection of MCI and dementia based on the Chinese Neuropsychological Consensus Battery (CNCB) in the Chinese Neuropsychological Normative Project (CN-NORM) cohort. METHODS CN-NORM was a nationwide, multicenter study conducted in China with 871 participants, including an MCI group (n=327, 37.5%), a dementia group (n=186, 21.4%), and a cognitively normal (CN) group (n=358, 41.1%). We used the following 4 algorithms to select candidate variables: the <i>F</i>-score according to the SelectKBest method, the area under the curve (AUC) from logistic regression (LR), <i>P</i> values from the logit method, and backward stepwise elimination. Different models were constructed after considering the administration duration and complexity of combinations of various tests. Receiver operating characteristic curve and AUC metrics were used to evaluate the discriminative ability of the models via stratified sampling cross-validation and LR and support vector classification (SVC) algorithms. This model was further validated in the Alzheimer’s Disease Neuroimaging Initiative phase 3 (ADNI-3) cohort (N=743), which included 416 (56%) CN subjects, 237 (31.9%) patients with MCI, and 90 (12.1%) patients with dementia. RESULTS Except for social cognition, all other domains in the CNCB differed between the MCI and CN groups (<i>P</i><.008). In feature selection results regarding discrimination between the MCI and CN groups, the Hopkins Verbal Learning Test-5 minutes Recall had the best performance, with the highest mean AUC of up to 0.80 (SD 0.02) and an <i>F</i>-score of up to 258.70. The scalability of model 5 (Hopkins Verbal Learning Test-5 minutes Recall and Trail Making Test-B) was the lowest. Model 5 achieved a higher level of discrimination than the Hong Kong Brief Cognitive test score in distinguishing between the MCI and CN groups (<i>P</i><.05). Model 5 also provided the highest sensitivity of up to 0.82 (range 0.72-0.92) and 0.83 (range 0.75-0.91) according to LR and SVC, respectively. This model yielded a similar robust discriminative performance in the ADNI-3 cohort regarding differentiation between the MCI and CN groups, with a mean AUC of up to 0.81 (SD 0) according to both LR and SVC algorithms. CONCLUSIONS We developed a stable and scalable composite neurocognitive test based on ML that could differentiate not only between patients with MCI and controls but also between patients with different stages of cognitive impairment. This composite neurocognitive test is a feasible and practical digital biomarker that can potentially be used in large-scale cognitive screening and intervention studies.
Purpose: This study sought to examine associations between depression and unhealthy lifestyle behaviors in Chinese patients with acute coronary syndromes (ACS). Methods: This cross-sectional study included 4043 ACS patients from 16 hospitals across China who participated in the I-Care (Integrating Depression Care in Acute Coronary Syndromes Patients) trial. Patients were enrolled between November 2014 and January 2017. Depression was assessed with the Patient Health Questionnaire-9 (PHQ-9). Five lifestyle behaviors were assessed: smoking, drinking, body mass index (BMI), physical activity, and sleep quality. Results: A total of 135 patients (3.3%) were considered clinically depressed (PHQ-9 ≥10). After adjusting for covariates, physical activity and sleep quality were inversely related to PHQ-9 scores. Adjusted logistic models showed that depressed patients were 1.7 times likely to be physically inactive (OR = 1.74; 95% CI, 1.15-2.64) and 4.6 times likely to have poor sleep quality (OR = 4.60; 95% CI, 3.07-6.88) compared with nondepressed patients. The association of depression with smoking, unhealthy drinking, and unhealthy BMI was not significant after adjustment for demographic characteristics. Higher depression scores were found to be associated with a greater number of unhealthy lifestyle behaviors ( P for trend < .001). Conclusions: The association of depression and unhealthy lifestyles in post-ACS patients suggests that reducing depressive symptoms and improving healthy lifestyle behaviors could potentially improve clinical outcomes in this vulnerable patient population.
Objectives Psychological pain may be helpful in conceptualizing suicidal behavior, in that high motivation to avoid pain combined with painful feelings may contribute to an increased risk of suicide. However, no experimental study has tested this hypothesis. The aim of the present study is to provide empirical evidence for the relationship between anhedonia, pain avoidance motivation, and suicidal ideation. Method The sample comprised 40 depressed outpatients and 20 healthy control subjects. All participants completed the Beck Scale for Suicide Ideation (BSS), Beck Depression Inventory, Psychache Scale, Three‐Dimensional Psychological Pain Scale, the monetary incentive delay (MID), and affective incentive delay (AID) tasks. Based on BSS scores, clinical participants were divided into high suicidal ideation (HSI) and low suicidal ideation (LSI) groups. Results In the AID task, the HSI group had longer response times (RTs) under the reward condition than those under the punishment condition ( p = .002). The LSI and control groups had shorter RTs under the reward condition compared with those under the neural condition ( p <.001 and p = .008, respectively). The LSI group also had shorter RTs under the reward condition than under the punishment condition ( p = .003). Pain arousal ( r = −.33, p <.01) and BSS scores were significantly negatively correlated with differences in RTs between neutral and reward conditions. Pain avoidance ( r = .35, p <.01) and BSS scores were positively correlated with differences in RTs between neutral and punishment conditions. Conclusions The AID task was more sensitive than the MID task for the detection of participants’ motivation in approaching hedonic experiences and avoiding pain. A suicidal mindset is manifested as decreased motivation to experience hedonia and increased motivation to avoid pain, which could be strong predictors of suicidal behavior.
Juxtaglomerular cell tumor (JGCT) is a rare tumor, with approximately 100 cases reported in the literature. The authors respectively studied the clinical data of 11 patients diagnosed with JGCT in Peking Union Medical College Hospital from 2004 to 2014, and investigated the immunohistochemical profiles in 10 tumors. Nine of the 11 patients were diagnosed before the age of 40 years. Hypertension was present in all patients, while hypokalemia occurred in seven of 11 patients. Computed tomography detected JGCTs with a sensitivity of 100%. Immunoreactivities for CD34 and vascular endothelial growth factor were observed in most tumor specimens, suggesting that JGCTs express a variety of vessel-related immunohistochemical markers, although JGCTs are considered a tumor without abundant blood supply. Nuclear accumulation of cyclin D1 was common in JGCTs. Results from immunohistochemistry were negative for BRAF, HER2, and TFE3, suggesting that BRAF, HER2, and TFE3 genes might not play a part in tumorigenesis in JGCTs.
Low-Rank Adaptation (LoRA) is a parameter-efficient fine-tuning method that has been widely adopted in various downstream applications of LLMs. Together with the Mixture-of-Expert (MoE) technique, fine-tuning approaches have shown remarkable improvements in model capability. However, the coordination of multiple experts in existing studies solely relies on the weights assigned by the simple router function. Lack of communication and collaboration among experts exacerbate the instability of LLMs due to the imbalance load problem of MoE. To address this issue, we propose a novel MoE graph-based LLM fine-tuning framework GraphLoRA, in which a graph router function is designed to capture the collaboration signals among experts by graph neural networks (GNNs). GraphLoRA enables all experts to understand input knowledge and share information from neighbor experts by aggregating operations. Besides, to enhance each expert's capability and their collaborations, we design two novel coordination strategies: the Poisson distribution-based distinction strategy and the Normal distribution-based load balance strategy. Extensive experiments on four real-world datasets demonstrate the effectiveness of our GraphLoRA in parameter-efficient fine-tuning of LLMs, showing the benefits of facilitating collaborations of multiple experts in the graph router of GraphLoRA.
Age, education, and gender are the most common covariates used to define normative standards against which neuropsychological (NP) performance is interpreted, but influences of other demographic factors have begun to be appreciated. In developing nations, urban versus rural residence may differentially affect numerous factors that could influence cognitive test performances, including quality of both formal and informal educational experiences and employment opportunities. Such disparities may necessitate corrections for urban/rural (U/R) status in NP norms. Prior investigations of the U/R effect on NP performance typically have been confounded by differences in educational attainment. We addressed in this by comparing the NP performance of large, Chinese urban (Yunnan Province, n = 201) and rural (Anhui Province, n = 141) cohorts of healthy adults, while controlling for other demographic differences. Although the groups did not differ in global NP scores, a more complex pattern was observed within specific NP ability domains and tests. Urban participants showed better performance in select measures of processing speed and executive functions, verbal fluency, and verbal learning. Self-reported daily use of academic skills was predictive of many U/R differences. Controlling for academic skill use abrogated most U/R differences but revealed rural advantages in select measures of visual reasoning and motor dexterity. (JINS, 2011, 17, 000–000)
The purpose of this study was to evaluate the application of the minimum clinically important difference (MCID) concept to clinical results in Chinese patients with acutely exacerbated schizophrenia. The original study was an 8-week, open-label, single-arm, multicenter study of flexible doses of paliperidone-extended release (pali-ER) in Chinese patients with acutely exacerbated schizophrenia. This is a post hoc analysis to determine the MCID value of PANSS, PSP and evaluate the responsiveness of each outcome measurements in the acute phase of schizophrenia. The responsiveness of the four measurements (PANSS, PANSS reduction rate, PSP, CGI-S) was analyzed. Four hundred ninety nine patients completed the 8-week follow-up and were finally used for this post hoc analysis. The MCID calculated by different approaches varied from 14.02 to 31.50 for PANSS, 15.14 to 42.79% for PANSS reduction rate, and 7.62 to 13.13% for PSP. In addition, the improvement of the CGI-S owned the highest responsiveness of the four outcome measurements. The threshold value of MCID for schizophrenia patients was determined by choice of the assessment method to an extent. In addition, the CGI-S score appeared to be the most valid and responsive measure of effectiveness for the acute phase of schizophrenia when take the treatment satisfaction of patients as anchor.
Abstract Background: Fatigue is one of the most prevalent and debilitating symptoms of major depressive disorder (MDD). The effective management of depression-related fatigue has an important impact on the patient's abilities, functioning, and quality of life (QOL). Moxibustion has been widely used in Traditional Chinese Medicine to manage fatigue. Recent studies have also demonstrated that moxibustion is effective for treating cancer-related fatigue and chronic fatigue syndrome. However, there is not sufficient data supporting the effect of moxibustion for depression-related fatigue. Therefore, this randomized, assessor-blinded, wait-list controlled trial is designed to evaluate the effectiveness, safety, and feasibility of moxibustion treatment for depression-related fatigue. Methods: One hundred and seventy-six participants who meet the diagnostic criteria for depression in the International Classification of Diseases, tenth revision (ICD-10), and who also have a score of ≥1 on the 13 th item of the Hamilton Depression Rating Scale-17 (HAMD-17), will be enrolled. At study entry, participants will undergo anti-depressant treatment for at least 1 month. Then those who still have a score of ≥1 on the 13 th item of the HAMD-17 will be randomly allocated to either a moxibustion group or wait-list control group in a ratio of 1:1. Anti-depressants will be provided for both groups during the whole process of the study period. Participants in the moxibustion group will undergo 14 sessions of moxibustion (over 2 weeks) with anti-depressant treatment, and participants in the wait-list control group will receive only anti-depressant treatment. Subsequently, participants in the moxibustion group will be followed-up for 4 weeks. The primary outcome measure will be the Fatigue Severity Scale (FSS). The secondary outcome measure will be the HAMD-17. Safety will be assessed by monitoring adverse events during the study. Trial feasibility will also be assessed in this study. Discussion: The results of this study may provide evidence for the efficacy of moxibustion as an adjunct to antidepressants for depression-related fatigue, and promote a more widespread foundation for the selection of moxibustion in the clinical setting as well as for future research in moxibustion therapy. Trial registration: This study protocol was registered at the Chinese Clinical Trial Registry (ChiCTR1800016905).
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)