Although multiple studies have identified skipping breakfast as a risk factor for weight gain, there is limited evidence on the clinical impact of skipping lunch and dinner on weight gain. This retrospective cohort study including 17,573 male and 8860 female university students at a national university in Japan, assessed the association of the frequency of breakfast, lunch, and dinner with the incidence of weight gain (≥10%) and overweight/obesity (body mass index ≥ 25 kg/m²), using annual participant health checkup data. Within the observation period of 3.0 ± 0.9 years, the incidence of ≥10% weight gain was observed in 1896 (10.8%) men and 1518 (17.1%) women, respectively. Skipping dinner was identified as a significant predictor of weight gain in multivariable-adjusted Poisson regression models for both men and women (skipping ≥ occasionally vs. eating every day, adjusted incidence rate ratios, 1.42 (95% confidence interval: 1.02–1.98) and 1.67 (1.33–2.09) in male and female students, respectively), whereas skipping breakfast and lunch were not. Similarly, skipping dinner, not breakfast or lunch, was associated with overweight/obesity (1.74 (1.07–2.84) and 1.68 (1.02–2.78) in men and women, respectively). In conclusion, skipping dinner predicted the incidence of weight gain and overweight/obesity in university students.

Previous studies have reported conflicting results on the clinical impact of alcohol consumption on the glomerular filtration rate (GFR). This retrospective cohort study aimed to assess the dose-dependent association between alcohol consumption and the slope of the estimated GFR (eGFR) in 304,929 participants aged 40–74 years who underwent annual health checkups in Japan between April 2008 and March 2011. The association between the baseline alcohol consumption and eGFR slope during the median observational period of 1.9 years was assessed using linear mixed-effects models with the random intercept and random slope of time adjusting for clinically relevant factors. In men, rare drinkers and daily drinkers with alcohol consumptions of ≥60 g/day had a significantly larger decline in eGFR than occasional drinkers (difference in multivariable-adjusted eGFR slope with 95% confidence interval (mL/min/1.73 m²/year) of rare, occasional, and daily drinkers with ≤19, 20–39, 40–59, and ≥60 g/day: −0.33 [−0.57, −0.09], 0.00 [reference], −0.06 [−0.39, 0.26], −0.16 [−0.43, 0.12], −0.08 [−0.47, 0.30], and −0.79 [−1.40, −0.17], respectively). In women, only rare drinkers were associated with lower eGFR slopes than occasional drinkers. In conclusion, alcohol consumption was associated with the eGFR slope in an inverse U-shaped fashion in men but not in women.

DOI: https://doi.org/10.1103/PhysRevC.81.059901

Key Points: Self-supervised learning extracts meaningful glomerular features without teacher labels. Self-distillation with no labels–pretrained model outperformed conventional supervised learning in disease and clinical parameter classification. Self-distillation with no labels model enabled deep learning on small datasets, reducing annotation efforts. Background: Deep learning has great potential in digital kidney pathology. However, its effectiveness depends heavily on the availability of extensively labeled datasets, which are often limited because of the specialized knowledge and time required for their creation. This limitation hinders the widespread application of deep learning for the analysis of kidney biopsy images. Methods: We applied self-distillation with no labels (DINO), a self-supervised learning method, to a dataset of 10,423 glomerular images obtained from 384 periodic acid–Schiff-stained kidney biopsy slides. Glomerular features extracted from the DINO-pretrained backbone were visualized using principal component analysis. We then performed classification tasks by adding either k-nearest neighbor classifiers or linear head layers to the DINO-pretrained or ImageNet-pretrained backbones. These models were trained on our labeled classification dataset. Performance was evaluated using metrics such as the area under the receiver operating characteristic curve (ROC-AUC). The classification tasks encompassed four disease categories (minimal change disease, mesangial proliferative GN, membranous nephropathy, and diabetic nephropathy) and clinical parameters such as hypertension, proteinuria, and hematuria. Results: Principal component analysis visualization revealed distinct principal components corresponding to different glomerular structures, demonstrating the capability of the DINO-pretrained backbone to capture morphologic features. In disease classification, the DINO-pretrained transferred model (ROC-AUC=0.93) outperformed the ImageNet-pretrained fine-tuned model (ROC-AUC=0.89). When the labeled data were limited, the ImageNet-pretrained fine-tuned model's ROC-AUC dropped to 0.76 (95% confidence interval, 0.72 to 0.80), whereas the DINO-pretrained transferred model maintained superior performance (ROC-AUC, 0.88; 95% confidence interval, 0.86 to 0.90). The DINO-pretrained transferred model also exhibited higher AUCs for the classification of several clinical parameters. External validation using two independent datasets confirmed DINO pretraining's superiority, particularly when labeled data were limited. Conclusions: The application of DINO to unlabeled periodic acid–Schiff-stained glomerular images facilitated the extraction of histologic features that could be effectively used for disease classification.

Although multiple studies have revealed a close association of skipping breakfast with cardiometabolic diseases, few studies have reported its association with chronic kidney disease (CKD). Furthermore, there is scant reporting on the clinical impacts that skipping lunch and dinner has on cardiometabolic diseases and CKD. This retrospective cohort study, including 5439 female and 4674 male workers of a national university in Japan who underwent annual health checkups between January 2005 and March 2013, aimed to assess an association of frequencies of breakfast, lunch, and dinner with incidence of proteinuria (dipstick urinary protein ≥1+). The incidence of proteinuria was observed in 763 (14.0%) females and 617 (13.2%) males during the median 4.3 and 5.9 years of the observational period, respectively. In females, skipping breakfast as well as skipping dinner, but not lunch, were associated with the incidence of proteinuria (adjusted hazard ratios of breakfast frequency of “every day”, “sometimes”, and “rarely”: 1.00 (reference), 1.35 (1.09–1.66), and 1.54 (1.22–1.94), respectively; those of dinner frequency of “every day” and “≤sometimes”: 1.00 (reference) and 1.31 (1.00–1.72), respectively). However, no association was observed in male workers. Skipping breakfast and skipping dinner were identified as risk factors of proteinuria in females, but not in males.

Introduction: Little information is available about the association between vegetable preference and chronic kidney disease. Methods: This retrospective cohort study included 10,819 university workers in Japan who underwent their annual health checkups between January 2005 and March 2013. According to a question “Do you like vegetables”? with 3 possible answers of “I like vegetables,” “I like vegetables somewhat,” or “I dislike vegetables,” 2,831, 2,249, and 104 male workers and 3,902, 1,648, and 85 female workers were classified into the “like,” “somewhat,” and “dislike” groups, respectively. An association between vegetable preference and incidence of proteinuria (dipstick urinary protein ≥1+) was assessed using Cox proportional-hazards models adjusted for clinically relevant factors. Results: During the median observational period of 5.0 years, the incidence of proteinuria was observed in 650 (12.7%) male and 789 (14.1%) female workers. Among male workers, the “dislike” group had a significantly higher risk of proteinuria (multivariable-adjusted hazard ratio of “like,” “somewhat,” and “dislike” groups: 1.00 [reference], 1.05 [0.90–1.23], and 1.59 [1.01–2.50], respectively). Among female workers, vegetable preference was associated with the incidence of proteinuria in a dose-dependent manner (1.00 [reference], 1.20 [1.04–1.40], 1.95 [1.26–3.02], respectively). Conclusion: “Do you like vegetables”? was a clinically useful tool to identify subjects vulnerable to proteinuria.