ＨＥ ＹＵＰＥＮＧ

Aim Japan faces a public health challenge of dementia, further complicated by the increasing complications from diabetes within its rapidly aging population. This study assesses the impact of diabetes on mortality and hospitalization among individuals aged ≥75 years with new dementia diagnoses. Methods We analyzed administrative claims data in Japan from 73 324 individuals aged ≥75 years with dementia, of whom 17% had comorbid diabetes. Dementia and diabetes were identified from the International Classification of Diseases, Tenth Revision codes. We used Kaplan–Meier survival analysis, Cox proportional hazards analysis, and population attributable fractions (PAFs) to evaluate the impact on mortality and hospitalization after dementia diagnosis. Results One‐year mortality and 1‐year hospitalization probabilities in individuals with dementia and diabetes (10.3% and 31.7%, respectively) were higher than those without diabetes (8.3% and 25.4%, respectively). The adjusted hazard ratios for individuals with diabetes, as compared to those without, were 1.126 (95% confidence interval [CI], 1.040–1.220) for mortality and 1.191 (95% CI, 1.140–1.245) for hospitalization. The PAFs from the comorbidity of dementia and diabetes were 2.2% for mortality and 3.1% for hospitalization. Subgroup analysis showed that the PAFs were highest in men aged 75–79 years and women aged 80–84 years for mortality and in individuals aged 75–79 for hospitalization. Conclusion During the early postdiagnosis period, comorbid diabetes increases mortality and hospitalization risks in older adults with dementia. The variation in disease burden across age groups underscores the need for age‐specific health care strategies to manage comorbid diabetes in individuals with dementia. Geriatr Gerontol Int 2024; 24: 773–781.

Aim Several studies have shown that dairy consumption in old age is effective in preventing frailty. However, there is a lack of evidence regarding the association between milk consumption during middle age and the development of frailty in old age. Therefore, we carried out an investigation to explore the association between milk consumption during middle age and development of frailty examined after over 15 years of follow up in a long‐term cohort study in Japan. Methods We studied 265 participants aged 60–79 years (212 men and 53 women) in 2018, who participated in both the baseline survey in 2002 and the frailty assessment in 2018. The amount of milk consumption (g/day) at baseline was age‐ and energy‐adjusted, and classified into three categories (no, low and high consumption: 0 g/day, ≤135.86 g/day, >135.86 g/day in men and 0 g/day, ≤126.44 g/day, >126.44 g/day in women). Odds ratios (OR) and 95% confidence intervals (CI) for prefrailty/frailty after adjusting for lifestyles at baseline, stratified by sex, were estimated using logistic regression analysis. Results The prevalence of prefrailty/frailty in 2018 was 37.7% and 28.3% in men and women, respectively. Milk consumption categories were inversely associated with the prevalence of prefrailty/frailty in men (OR 0.34, 95% CI 0.14–0.84 in low consumption; OR 0.31, 95% CI 0.10–0.95 in high consumption; P < 0.05), but not in women (OR 0.53, 95% CI 0.11–2.65; P = 0.44). Conclusions In this study, milk intake in middle‐aged men was inversely associated with the prevalence of prefrailty/frailty later in life. Geriatr Gerontol Int 2024; 24: 700–705.

Background and Objective: Excellent generalizability is the precondition for the widespread practical implementation of machine learning models. In our previous study, we developed the schizophrenia classification model (SZ classifier) to identify potential schizophrenia patients in the Japanese population. The SZ classifier has exhibited impressive performance during internal validation. However, ensuring the robustness and generalizability of the SZ classifier requires external validation across independent sample sets. In this study, we aimed to present an external validation of the SZ classifier using outpatient data. Methods: The SZ classifier was trained by using online survey data, which incorporate demographic, health-related, and social comorbidity features. External validation was conducted using an outpatient sample set which is independent from the sample set during the model development phase. The model performance was assessed based on the sensitivity and misclassification rates for schizophrenia, bipolar disorder, and major depression patients. Results: The SZ classifier demonstrated a sensitivity of 0.75 when applied to schizophrenia patients. The misclassification rates were 59% and 55% for bipolar disorder and major depression patients, respectively. Conclusions: The SZ classifier currently encounters challenges in accurately determining the presence or absence of schizophrenia at the individual level. Prior to widespread practical implementation, enhancements are necessary to bolster the accuracy and diminish the misclassification rates. Despite the current limitations of the model, such as poor specificity for certain psychiatric disorders, there is potential for improvement if including multiple types of psychiatric disorders during model development.

Abstract Objectives This study aimed to develop an approach to enhance the model precision by artificial images. Materials and Methods Given an epidemiological study designed to predict 1 response using f features with M samples, each feature was converted into a pixel with certain value. Permutated these pixels into F orders, resulting in F distinct artificial image sample sets. Based on the experience of image recognition techniques, appropriate training images results in higher precision model. In the preliminary experiment, a binary response was predicted by 76 features, the sample set included 223 patients and 1776 healthy controls. Results We randomly selected 10 000 artificial sample sets to train the model. Models’ performance (area under the receiver operating characteristic curve values) depicted a bell-shaped distribution. Conclusion The model construction strategy developed in the research has potential to capture feature order related information and enhance model predictability.