Assessing the healthiness of menus of all out-of-home food outlets and its socioeconomic patterns in Great Britain.

Food environment research predominantly focuses on the spatial distribution of out-of-home food outlets. However, the healthiness of food choices available within these outlets has been understudied, largely due to resource constraints. In this study, we propose an innovative, low-resource approach to characterise the healthiness of out-of-home food outlets at scale. Menu healthiness scores were calculated for food outlets on JustEat, and a deep learning model was trained to predict these scores for all physical out-of-home outlets in Great Britain, based on outlet names. Our findings highlight the "double burden" of the unhealthy food environment in deprived areas where there tend to be more out-of-home food outlets, and these outlets tend to be less healthy. This methodological advancement provides a nuanced understanding of out-of-home food environments, with potential for automation and broad geographic application.

dsSurvival 2.0: privacy enhancing survival curves for survival models in the federated DataSHIELD analysis system.

OBJECTIVE: Survival models are used extensively in biomedical sciences, where they allow the investigation of the effect of exposures on health outcomes. It is desirable to use diverse data sets in survival analyses, because this offers increased statistical power and generalisability of results. However, there are often challenges with bringing data together in one location or following an analysis plan and sharing results. DataSHIELD is an analysis platform that helps users to overcome these ethical, governance and process difficulties. It allows users to analyse data remotely, using functions that are built to restrict access to the detailed data items (federated analysis). Previous works have provided survival modelling functionality in DataSHIELD (dsSurvival package), but there is a requirement to provide functions that offer privacy enhancing survival curves that retain useful information. RESULTS: We introduce an enhanced version of the dsSurvival package which offers privacy enhancing survival curves for DataSHIELD. Different methods for enhancing privacy were evaluated for their effectiveness in enhancing privacy while maintaining utility. We demonstrated how our selected method could enhance privacy in different scenarios using real survival data. The details of how DataSHIELD can be used to generate survival curves can be found in the associated tutorial.

Monitoring the Nutrient Composition of Food Prepared Out-of-Home in the United Kingdom: Database Development and Case Study.

BACKGROUND: Hand transcribing nutrient composition data from websites requires extensive human resources and is prone to error. As a result, there are limited nutrient composition data on food prepared out of the home in the United Kingdom. Such data are crucial for understanding and monitoring the out-of-home food environment, which aids policy making. Automated data collection from publicly available sources offers a potential low-resource solution to address this gap. OBJECTIVE: In this paper, we describe the first UK longitudinal nutritional database of food prepared out of the home, MenuTracker. As large chains will be required to display calorie information on their UK menus from April 2022, we also aimed to identify which chains reported their nutritional information online in November 2021. In a case study to demonstrate the utility of MenuTracker, we estimated the proportions of menu items exceeding recommended energy and nutrient intake (eg, >600 kcal per meal). METHODS: We have collated nutrient composition data of menu items sold by large chain restaurants quarterly since March 2021. Large chains were defined as those with 250 employees or more (those covered by the new calorie labeling policy) or belonging to the top 100 restaurants based on sales volume. We developed scripts in Python to automate the data collection process from business websites. Various techniques were used to harvest web data and extract data from nutritional tables in PDF format. RESULTS: Automated Python programs reduced approximately 85% of manual work, totaling 500 hours saved for each wave of data collection. As of January 2022, MenuTracker has 76,405 records from 88 large out-of-home food chains at 4 different time points (ie, March, June, September, and December) in 2021. In constructing the database, we found that one-quarter (24.5%, 256/1043) of large chains, which are likely to be subject to the United Kingdom's calorie menu labeling regulations, provided their nutritional information online in November 2021. Across these chains, 24.7% (16,391/66,295) of menu items exceeded the UK government's recommendation of a maximum of 600 kcal for a single meal. Comparable figures were 46.4% (29,411/63,416) for saturated fat, 34.7% (21,964/63,388) for total fat, 17.6% (11,260/64,051) for carbohydrates, 17.8% (11,434/64,059) for sugar, and 35.2% (22,588/64,086) for salt. Furthermore, 0.7% to 7.1% of the menu items exceeded the maximum daily recommended intake for these nutrients. CONCLUSIONS: MenuTracker is a valuable resource that harnesses the power of data science techniques to use publicly available data online. Researchers, policy makers, and consumers can use MenuTracker to understand and assess foods available from out-of-home food outlets. The methods used in development are available online and can be used to establish similar databases elsewhere.

dsSurvival: Privacy preserving survival models for federated individual patient meta-analysis in DataSHIELD.

OBJECTIVE: Achieving sufficient statistical power in a survival analysis usually requires large amounts of data from different sites. Sensitivity of individual-level data, ethical and practical considerations regarding data sharing across institutions could be a potential challenge for achieving this added power. Hence we implemented a federated meta-analysis approach of survival models in DataSHIELD, where only anonymous aggregated data are shared across institutions, while simultaneously allowing for exploratory, interactive modelling. In this case, meta-analysis techniques to combine analysis results from each site are a solution, but an analytic workflow involving local analysis undertaken at individual studies hinders exploration. Thus, the aim is to provide a framework for performing meta-analysis of Cox regression models across institutions without manual analysis steps for the data providers. RESULTS: We introduce a package (dsSurvival) which allows privacy preserving meta-analysis of survival models, including the calculation of hazard ratios. Our tool can be of great use in biomedical research where there is a need for building survival models and there are privacy concerns about sharing data.

dsSynthetic: synthetic data generation for the DataSHIELD federated analysis system.

OBJECTIVE: Platforms such as DataSHIELD allow users to analyse sensitive data remotely, without having full access to the detailed data items (federated analysis). While this feature helps to overcome difficulties with data sharing, it can make it challenging to write code without full visibility of the data. One solution is to generate realistic, non-disclosive synthetic data that can be transferred to the analyst so they can perfect their code without the access limitation. When this process is complete, they can run the code on the real data. RESULTS: We have created a package in DataSHIELD (dsSynthetic) which allows generation of realistic synthetic data, building on existing packages. In our paper and accompanying tutorial we demonstrate how the use of synthetic data generated with our package can help DataSHIELD users with tasks such as writing analysis scripts and harmonising data to common scales and measures.

Associations between exploratory dietary patterns and incident type 2 diabetes: a federated meta-analysis of individual participant data from 25 cohort studies.

PURPOSE: In several studies, exploratory dietary patterns (DP), derived by principal component analysis, were inversely or positively associated with incident type 2 diabetes (T2D). However, findings remained study-specific, inconsistent and rarely replicated. This study aimed to investigate the associations between DPs and T2D in multiple cohorts across the world. METHODS: This federated meta-analysis of individual participant data was based on 25 prospective cohort studies from 5 continents including a total of 390,664 participants with a follow-up for T2D (3.8-25.0 years). After data harmonization across cohorts we evaluated 15 previously identified T2D-related DPs for association with incident T2D estimating pooled incidence rate ratios (IRR) and confidence intervals (CI) by Piecewise Poisson regression and random-effects meta-analysis. RESULTS: 29,386 participants developed T2D during follow-up. Five DPs, characterized by higher intake of red meat, processed meat, French fries and refined grains, were associated with higher incidence of T2D. The strongest association was observed for a DP comprising these food groups besides others (IRRpooled per 1 SD = 1.104, 95% CI 1.059-1.151). Although heterogeneity was present (I2 = 85%), IRR exceeded 1 in 18 of the 20 meta-analyzed studies. Original DPs associated with lower T2D risk were not confirmed. Instead, a healthy DP (HDP1) was associated with higher T2D risk (IRRpooled per 1 SD = 1.057, 95% CI 1.027-1.088). CONCLUSION: Our findings from various cohorts revealed positive associations for several DPs, characterized by higher intake of red meat, processed meat, French fries and refined grains, adding to the evidence-base that links DPs to higher T2D risk. However, no inverse DP-T2D associations were confirmed.

Socioeconomic inequalities in food outlet access through an online food delivery service in England: A cross-sectional descriptive analysis.

Online food delivery services facilitate 'online' access to food outlets selling food prepared away-from-home. Online food outlet access has not previously been investigated in England or across an entire country. Systematic differences in online food outlet access could exacerbate existing health inequalities, which is a public health concern. However, this is not known. Across postcode districts in England (n = 2118), we identified and described the number of food outlets and unique cuisine types accessible online from the market leader (Just Eat). We investigated associations with area-level deprivation using adjusted negative binomial regression models. We also compared the number of food outlets accessible online with the number physically accessible in the neighbourhood (1600m Euclidean buffers of postcode district geographic centroids) and investigated associations with deprivation using an adjusted general linear model. For each outcome, we predicted means and 95% confidence intervals. In November 2019, 29,232 food outlets were registered to accept orders online. Overall, the median number of food outlets accessible online per postcode district was 63.5 (IQR; 16.0-156.0). For the number of food outlets accessible online as a percentage of the number accessible within the neighbourhood, the median was 63.4% (IQR; 35.6-96.5). Analysis using negative binomial regression showed that online food outlet access was highest in the most deprived postcode districts (n = 106.1; 95% CI: 91.9, 120.3). The number of food outlets accessible online as a percentage of those accessible within the neighbourhood was highest in the least deprived postcode districts (n = 86.2%; 95% CI: 78.6, 93.7). In England, online food outlet access is socioeconomically patterned. Further research is required to understand how online food outlet access is related to using online food delivery services.

Associations of Total Legume, Pulse, and Soy Consumption with Incident Type 2 Diabetes: Federated Meta-Analysis of 27 Studies from Diverse World Regions.

BACKGROUND: The consumption of legumes is promoted as part of a healthy diet in many countries but associations of total and types of legume consumption with type 2 diabetes (T2D) are not well established. Analyses across diverse populations are lacking despite the availability of unpublished legume consumption data in prospective cohort studies. OBJECTIVE: To examine the prospective associations of total and types of legume intake with the risk of incident T2D. METHODS: Meta-analyses of associations between total legume, pulse, and soy consumption and T2D were conducted using a federated approach without physical data-pooling. Prospective cohorts were included if legume exposure and T2D outcome data were available and the cohort investigators agreed to participate. We estimated incidence rate ratios (IRRs) and CIs of associations using individual participant data including ≤42,473 incident cases among 807,785 adults without diabetes in 27 cohorts across the Americas, Eastern Mediterranean, Europe, and Western Pacific. Random-effects meta-analysis was used to combine effect estimates and estimate heterogeneity. RESULTS: Median total legume intake ranged from 0-140 g/d across cohorts. We observed a weak positive association between total legume consumption and T2D (IRR = 1.02, 95% CI: 1.01 to 1.04) per 20 g/d higher intake, with moderately high heterogeneity (I2 = 74%). Analysis by region showed no evidence of associations in the Americas, Eastern Mediterranean, and Western Pacific. The positive association in Europe (IRR = 1.05, 95% CI: 1.01 to 1.10, I2 = 82%) was mainly driven by studies from Germany, UK, and Sweden. No evidence of associations was observed for the consumption of pulses or soy. CONCLUSIONS: These findings suggest no evidence of an association of legume intakes with T2D in several world regions. The positive association observed in some European studies warrants further investigation relating to overall dietary contexts in which legumes are consumed, including accompanying foods which may be positively associated with T2D.

Automatic classification of takeaway food outlet cuisine type using machine (deep) learning.

BACKGROUND AND PURPOSE: Researchers have not disaggregated neighbourhood exposure to takeaway ('fast-') food outlets by cuisine type sold, which would otherwise permit examination of differential impacts on diet, obesity and related disease. This is partly due to the substantial resource challenge of manual classification of unclassified takeaway outlets at scale. We describe the development of a new model to automatically classify takeaway food outlets, by 10 major cuisine types, based on business name alone. MATERIAL AND METHODS: We used machine (deep) learning, and specifically a Long Short Term Memory variant of a Recurrent Neural Network, to develop a predictive model trained on labelled outlets (n = 14,145), from an online takeaway food ordering platform. We validated the accuracy of predictions on unseen labelled outlets (n = 4,000) from the same source. RESULTS: Although accuracy of prediction varied by cuisine type, overall the model (or 'classifier') made a correct prediction approximately three out of four times. We demonstrated the potential of the classifier to public health researchers and for surveillance to support decision-making, through using it to characterise nearly 55,000 takeaway food outlets in England by cuisine type, for the first time. CONCLUSIONS: Although imperfect, we successfully developed a model to classify takeaway food outlets, by 10 major cuisine types, from business name alone, using innovative data science methods. We have made the model available for use elsewhere by others, including in other contexts and to characterise other types of food outlets, and for further development.