BackgroundObesity and weight gain is a critical public health concern. Serious digital games are gaining popularity in the context of health interventions. They use persuasive and fun design features to engage users in health-related behaviors in a non-game context. As a young field, research about effectiveness and acceptability of such games for weight loss is sparse.ObjectiveThe goal of this study was to evaluate real-world play patterns of SpaPlay and its impact on body mass index (BMI) and nutritional knowledge. SpaPlay is a computer game designed to help women adopt healthier dietary and exercise behaviors, developed based on Self-Determination theory and the Player Experience of Need Satisfaction (PENS) model. Progress in the game is tied to real-life activities (e.g., eating a healthy snack, taking a flight of stairs).MethodsWe recruited 47 women to partake in a within-subject 90-day longitudinal study, with assessments taken at baseline, 1-, 2-, and 3- months. Women were on average, 29.8 years old (±7.3), highly educated (80.9% had BA or higher), 39% non-White, baseline BMI 26.98 (±5.6), who reported at least contemplating making changes in their diet and exercise routine based on the Stages of Change Model. We computed 9 indices from game utilization data to evaluate game play. We used general linear models to examine inter-individual differences between levels of play, and multilevel models to assess temporal changes in BMI and nutritional knowledge.ResultsPatterns of game play were mixed. Participants who reported being in the preparation or action stages of behavior change exhibited more days of play and more play regularity compared to those who were in the contemplation stage. Additionally, women who reported playing video games 1-2 hours per session demonstrated more sparse game play. Brief activities, such as one-time actions related to physical activity or healthy food, were preferred over activities that require a longer commitment (e.g., taking stairs every day for a week). BMI decreased significantly (P<.001) from baseline to 3-month follow-up, yielding a large effect size of 1.28. Nutritional knowledge increased significantly (P<.001) from first to third month follow-ups, with an effect size of .86. The degree of change in both outcomes was related to game play, baseline readiness to change, and the extent of video game play in general.ConclusionsThis work demonstrates initial evidence of success for using a serious game as an intervention for health behavior change in real world settings. Our findings also highlight the need to understand not only game effectiveness but also inter-individual differences. Individualizing content and the intervention medium appears to be necessary for a more personalized and long-lasting impact.

Typical diets include an assortment of unprocessed, processed, and ultra-processed foods, along with culinary ingredients. Linear programming (LP) can be used to generate nutritionally adequate food patterns that meet pre-defined nutrient guidelines. The present LP models were set to satisfy 22 nutrient standards, while minimizing deviation from the mean observed diet of the Seattle Obesity Study (SOS III) sample. Component foods from the Fred Hutch food frequency questionnaire comprised the market basket. LP models generated optimized 2000 kcal food patterns by selecting from all foods, unprocessed foods only, ultra-processed foods only, or some other combination. Optimized patterns created using all foods contained less fat, sugar, and salt, and more vegetables compared to the SOS III mean. Ultra-processed foods were the main sources of added sugar, saturated fat and sodium. Ultra-processed foods also contributed most vitamin E, thiamin, niacin, folate, and calcium, and were the main sources of plant protein. LP models failed to create optimal diets using unprocessed foods only and ultra-processed foods only: no mathematical solution was obtained. Relaxing the vitamin D criterion led to optimized diets based on unprocessed or ultra-processed foods only. However, food patterns created using unprocessed foods were significantly more expensive compared to those created using foods in the ultra-processed category. This work demonstrates that foods from all NOVA categories can contribute to a nutritionally adequate diet.

No abstract

No abstract

Text detoxification has the potential to mitigate the harms of toxicity by rephrasing text to remove offensive meaning, but subtle toxicity remains challenging to tackle. We introduce MARCO, a detoxification algorithm that combines controllable generation and text rewriting methods using a Product of Experts with autoencoder language models (LMs). MARCO uses likelihoods under a non-toxic LM (expert) and a toxic LM (anti-expert) to find candidate words to mask and replace. We evaluate our method on several subtle toxicity and microaggressions datasets, and show that it not only outperforms baselines on automatic metrics, but MARCO's rewrites are preferred 2.1× more in human evaluation. Its applicability to instances of subtle toxicity is especially promising, demonstrating a path forward for addressing increasingly elusive online hate.the quality of online conversations (e.g., through machine-in-the-loop interfaces; Hohenstein et al., 2021;Clark et al., 2018).We present MARCO, Mask and Replace with Context: a new, unsupervised algorithm for text detoxification that combines mask-and-replace text