David B. Allison

Circadian biology is recognized for its importance to physiology and metabolism, with implications for healthy aging and disease risk. To test whether altering the number of imposed environmental cycles experienced per unit chronological time affects health biomarkers and lifespan, we compressed and expanded the photic cycle (T-cycle, combined light:dark length) by ~ 10% relative to the 24-h day. Male C57BL/6 J mice ( n = 60 per group, 12 weeks old) were randomized to one of three T-cycles: T ~ 22-, T = 24-, or T ~ 26-h. Seasonal variations in light:dark proportions of the photoperiod from 8:16 to 16:8 and back over 365 imposed photic cycles were applied to the T = 24 condition and scaled (+ / − 10%) for each. Body weight and food intake were measured weekly under ad libitum feeding, with body composition, metabolic rate, and functional measures collected. Mortality was measured daily, and survival was compared by Cox and linear mixed models (LMM). T-cycle group differences for health biomarkers were assessed with LMM. Mice gained weight early in life across all three T-cycles, with significant effects of time, T-cycle, and time*T-cycle interactions ( p < 0.001, p < 0.001, p < 0.001; T ~ 22 quicker weight gain). Similar results were found for fat mass and lean mass, where significant ( p < 0.05) time, T-cycle, and time*T-cycle interactions were observed. There was no significant difference among T-cycles for overall survival, mean longevity, or maximum lifespan ( p = 0.086, p = 0.079, and p = 0.527, respectively). Mice experiencing 10% chronic contraction or extension of the daily T-cycle do not show significant differences in mortality versus the 24-h T-cycle controls.

The credibility of nutritional research is dependent on the rigor with which studies are conducted and the ability for independent assessment to be performed. Despite the importance of these, more work is needed in the field of nutrition to buttress the trustworthiness of nutrition research. To develop and apply a process for evaluating the rigor, reproducibility, and verifiability of nutritional research, using the relationship between potato consumption and Colorectal Cancer (CRC) as a case study. We updated existing systematic reviews to include studies on potatoes and CRC, assessing their design, execution, and reporting quality. We attempted to reproduce and verify the results of included studies by requesting raw data from authors and following statistical methods as described in the publications. Rigor was evaluated using four different tools: ROBINS-E, STROBE-Nut, Newcastle-Ottawa scale, and additional criteria related to transparency. Eighteen studies were included, none of which publicly share data. We managed to access data for only two studies, successfully reproducing and verifying the results for one. The majority of studies exhibited a high risk of bias, with significant limitations in reporting quality and methodological rigor. Research on the relationship between potato consumption and CRC risk is insufficiently reproducible and verifiable, undermining the trustworthiness of its findings. This study highlights the need for improving transparency, data sharing, and methodological rigor in nutritional research. Our approach provides a model for assessing the credibility of research in other areas of nutrition.

Pork is a common source of high-quality protein. However, it is typically recommended in moderation due to its saturated fat content, especially as its role in weight management, weight-loss (WL) programs, and overall diet quality is not well understood. This study aimed to identify associations among pork intake, WL, and diet quality among adults with overweight or obesity participating in a weight management program (WMP). This secondary analysis used data from a 6-month randomized controlled trial (NCT05648344) evaluating the effectiveness of a WMP on diet quality. Data were collected through online surveys and the Automated Self-Administered 24-h Dietary Assessment Tool, and diet quality was assessed with the Healthy-Eating Index-2020. Based on reported pork products at baseline, participants were categorized as pork consumers (PC) or no-pork consumers (NPC). ANCOVA assessed associations between pork intake status and percent WL; linear mixed models assessed pork intake status and diet quality; and logistic regressions evaluated the achievement of WL milestones by pork intake status. A total of 187 participants randomized to the WMP were included in the analysis. Participants' mean age: 47.6 ± 12.1 years; mean BMI: 33.6 ± 4.6 kg/m²; 5.9% Hispanic, 74.3% White; and 78.6% were PC. At 6 months, both PC and NPC achieved clinically meaningful WL, with PC averaging (SE) 5.3% (1.4), and NPC 8.6% (2.0). NPC attained significantly higher % WL, by mean (SE) 3.3 (1.4) % points, relative to PC, p = 0.019. Both groups experienced greater diet quality at 6 months, which increased by 4.3 (1.1) for PC, and by 2.5 (2.0) for NPC, p = 0.442. Habitual pork consumers in a WMP may achieve clinically meaningful WL and improve diet quality. However, more weight was lost among NPC, warranting further exploration.

The essentiality of protein in the human diet is unequivocal. Yet researchers, clinicians, and lay people often believe numerous propositions about dietary protein despite insufficient supporting or refuting data in some instances. To address this disconnect, and to "pressure-test" current beliefs about dietary protein, the Indiana University School of Public Health-Bloomington convened a workshop in February 2025 titled "Human Dietary Protein Needs and Benefits: A Critical Assessment of Postulated Propositions." More than 20 international experts discussed (1) methodologic issues affecting data acquisition and interpretation; (2) "optimal" dietary protein intakes and effects on muscle protein synthesis rates, muscle protein accretion, muscle growth, and muscle repair; (3) protein needs during weight loss; (4) acute protein intake thresholds above and below which protein is no longer related to anabolism; and (5) dietary protein intakes above which protein may be detrimental to health. The experts rated each proposition on a scale from "existing evidence strongly supports the proposition" to "existing evidence seems sufficient to rule out the viability of the proposition." In most instances, the experts believed additional research was warranted. For many propositions the research base was insufficient in terms of quality (rigor), quantity (sample size, study duration), or pertinence (e.g., use of surrogate markers).

Aging research relies on varied statistical methods, and applying these methods appropriately is important for scientific rigor. However, proper use of these statistical techniques is a challenge. We discuss two categories of statistical methods in aging research: ( a ) emerging methods requiring further validation, including techniques to examine compression of morbidity, maximum lifespan, immortal time bias, molecular aging clocks, and treatment response heterogeneity, and ( b ) classic and existing methods needing reconsideration and improvement, such as stepwise regression, generalized linear models, methods for accounting for clustering and nesting effects, methods for testing for group differences, methods for mediation and moderation analyses, and nonlinear models. For each method, we review its relevance to aging research, highlight statistical issues, and suggest improvements or alternatives with examples from aging research. We urge researchers to refine traditional approaches and embrace emerging methods tailored to the unique challenges of aging research. This review will help researchers identify and apply sound statistical methods, thereby improving statistical rigor in aging research.