Capturing children food exposure using wearable cameras and deep learning.

Children's dietary habits are influenced by complex factors within their home, school and neighborhood environments. Identifying such influencers and assessing their effects is traditionally based on self-reported data which can be prone to recall bias. We developed a culturally acceptable machine-learning-based data-collection system to objectively capture school-children's exposure to food (including food items, food advertisements, and food outlets) in two urban Arab centers: Greater Beirut, in Lebanon, and Greater Tunis, in Tunisia. Our machine-learning-based system consists of 1) a wearable camera that captures continuous footage of children's environment during a typical school day, 2) a machine learning model that automatically identifies images related to food from the collected data and discards any other footage, 3) a second machine learning model that classifies food-related images into images that contain actual food items, images that contain food advertisements, and images that contain food outlets, and 4) a third machine learning model that classifies images that contain food items into two classes, corresponding to whether the food items are being consumed by the child wearing the camera or whether they are consumed by others. This manuscript reports on a user-centered design study to assess the acceptability of using wearable cameras to capture food exposure among school children in Greater Beirut and Greater Tunis. We then describe how we trained our first machine learning model to detect food exposure images using data collected from the Web and utilizing the latest trends in deep learning for computer vision. Next, we describe how we trained our other machine learning models to classify food-related images into their respective categories using a combination of public data and data acquired via crowdsourcing. Finally, we describe how the different components of our system were packed together and deployed in a real-world case study and we report on its performance.

Prevalence of Hypertension and Adherence to Dietary Approaches to Stop Hypertension Diet Score in Childbearing Age Tunisian Women: A Cross-Sectional Study.

BACKGROUND AND AIMS: The prevalence, awareness, and treatment of hypertension, along with their sociodemographic, anthropometric, and lifestyle associations, were evaluated in a cross-sectional survey of childbearing age Tunisian women. Adherence to the Dietary Approaches to Stop Hypertension diet score was also assessed for hypertensive versus nonhypertensive women. METHODS AND RESULTS: A total of 1689 nonpregnant women, aged 20-49 years, were randomly sampled a regional (Greater Tunis), two-stage, stratified, cross-sectional cluster survey from March 2009 to January 2010. Data on medical history and sociodemographic characteristics were collected using a questionnaire. The average daily intake of energy and nutrients was computed using a specific Tunisian food composition database. The Dietary Approaches to Stop Hypertension diet score (0 to 10) was assessed by adding the individual scores (0 to 1) of ten nutrient components according to dietary guidelines. The overall prevalence of hypertension was 21.4%. Age, obesity, abdominal fat, parity, and family history were significantly associated with hypertension. The mean Dietary Approaches to Stop Hypertension accordance score was 4.93 for hypertensive women and 4.86 for nonhypertensive women (P = 0.0556). After adjustment for age, energy intake, and all nutritional covariates, no associations were observed between hypertension and Dietary Approaches to Stop Hypertension diet components. CONCLUSION: Though no clear-cut associations between hypertension and environmental or behavioral factors were identified in the study, the association with abdominal obesity and multiparity suggests that interventions aimed at lifestyle modifications to reduce these risk factors could be also useful in the prevention of hypertension.

Unbalanced intakes of sodium and potassium among Tunisian adults: A cross-sectional study.

The prevention and management of hypertension are untimely associated with a lowering of sodium intake. The present study aimed to evaluate the sodium and potassium intake levels of Tunisian population through measurement of 24-hr urinary sodium excretions. A randomly, multistage, cross-sectional study was conducted in an urban region (Bizerte) in Tunisia during 2015. The target population involved adults aged from 25 to 64 years. Sodium, potassium, and creatinine concentrations were determined in each urine sample using indirect potentiometric method. From the 420 selected participants, only 194 gave urine samples complying completeness criteria. A multivariate regression model was used to assess the variables related to sodium and potassium excretion. The daily mean excretion of sodium and potassium was 138.3 ± 46.5 mmol/d (corresponding to 8.1 ± 2.7 g/d of salt intake) and 61.0 ± 22.7 mmol/d, respectively. More than 87.1% of the participants (89.8% for men vs. 84.9% for women; p = .31) exceeded the WHO recommendation of 5 g/d. The upper limit of 10 g salt intake per day was still exceeded by 26.3%. After adjusted analysis, sex (for women, coef = -1.6; (95% CI: -2.4, -0.7)), level of instruction (≥30 kg/m2, coef = +1.1; (95% IC: 0.4-2.0)), and body mass index (≥30 kg/m2, coef = +1.1; (95% CI: 0.1, 2.0)) were associated with the sodium excretion. High sodium intake and inadequate potassium intake were found among participants. This consumption profile complies with the diet westernization context occurring in Tunisia. The initiated strategy focused on the downward of sodium in bread (the main source of salt intake) seems to be promising.

Gender inequalities in diet quality and their socioeconomic patterning in a nutrition transition context in the Middle East and North Africa: a cross-sectional study in Tunisia.

BACKGROUND: In a context of nutrition transition and major shifts in lifestyle and diet, the Middle East and North Africa features a marked gender excess adiposity gap detrimental to women. In this setting, where gender issues are especially acute, we investigated gender differences in dietary intake with a focus on diet quality, and how the differences varied with the area of residence and socio-demographic characteristics. METHODS: The study was conducted in 2009-2010 in the Greater Tunis region (Tunisia), as a case study of an advanced nutrition transition context in the region. A cross-sectional survey used a random, stratified, clustered sample of households: 1689 women and 930 men aged 20-49 years were analyzed. Dietary intake was assessed using a 3-day food record. Nutrient content was derived from a specific Tunisian food composition database. We analysed the Diet Quality Index-International (DQI-I) and sub-scores (variety, adequacy, moderation and balance). A score of DQI-I > 60 defined good diet quality. Inequality measures were women vs. men differences in means for interval variables and odds-ratios (OR) for DQI-I > 60. Their variation with socio-demographic characteristics was estimated using models featuring gender x covariate interactions. RESULTS: Mean energy intake/day was 2300 ± 15 kcal for women vs. 2859 ± 32 kcal for men. By 1000 g/kcal/d women consumed more fruits and sweets but less red meat and soft drinks than men. Women had a higher mean moderation sub-score than men (+ 1.8[1.4, 2.2], P < 0.0001) but lower variety (- 2.0[- 2.3, - 1.6], P < 0.0001) and adequacy (- 1.8[- 2.0, - 1.5], P < 0.0001). Thus, the overall mean DQI-I was lower among women than men (58.6 ± 0.3 vs. 60.4 ± 0.3, - 1.8[- 2.6, - 1.0], P < 0.0001) as was the proportion of DQI-I > 60 (45.2% vs. 55.7%, OR = 0.7[0.5, 0.8], P < 0.0001). Adjusted gender differences in DQI-I decreased with age but were higher in larger households and extreme categories of education (no-schooling and university) vs. the middle categories. CONCLUSION: In this nutrition transition context with only average diet quality, it was somewhat lower for women. Socioeconomic patterning of gender contrasts was mild. Beyond, that women had lower adequacy and variety scores but better moderation is a possible pathway for gender specific prevention messages.

Intra-household double burden of malnutrition in a North African nutrition transition context: magnitude and associated factors of child anaemia with mother excess adiposity.

OBJECTIVE: In the Middle East and North Africa region, the nutrition transition has resulted in drastic increases in excess adiposity, particularly among women, while some types of undernutrition remain prevalent, especially among pre-school children. We assessed the magnitude, nature and associated factors of the within-household co-occurrence of anaemia in children and excess adiposity in mothers. DESIGN: Cross-sectional survey using stratified two-stage random cluster sampling to survey households with women aged 20-49 years. BMI≥25·0 kg/m2 defined overweight and BMI≥30·0 kg/m2 obesity, while anaemia for children was defined as Hb<110 g/l. The associations between child anaemia and mother excess adiposity, and sociodemographic and lifestyle factors were estimated by multinomial regression. SETTING: Greater Tunis area, Tunisia, in 2009-2010. SUBJECTS: Children aged 6-59 months living with their 20-49-year-old mothers (437 child-mother pairs). RESULTS: The most prevalent double burden of malnutrition in child-mother pairs by far was the anaemic child and overweight mother (24·4 %; 95 % CI 20·1, 29·3 %). A significant proportion of pairs were anaemic child and obese mother (14·4 %; 95 % CI 11·0, 18·5 %). The co-occurrence of anaemia in child and excess adiposity in mother was neither synergetic nor antagonistic (P=0·59 and 0·40 for anaemia-overweight and anaemia-obesity, respectively). This double burden was more frequent among child-mother pairs with younger children, with mothers of higher parity and higher energy intakes. CONCLUSIONS: The high prevalence of anaemic child and overweight or obese mother requires special attention e.g. through interventions which simultaneously target both types of malnutrition within the same household.

Gender inequalities in excess adiposity and anaemia combine in a large double burden of malnutrition gap detrimental to women in an urban area in North Africa.

OBJECTIVE: The nutrition transition has exacerbated the gender gap in health in the Middle East and North Africa region as the increase in excess adiposity has been much higher among women than men. This is not exclusive of the persistence of anaemia, generally also more prevalent among women. We assessed the magnitude and sociodemographic factors associated with gender inequality vis-à-vis the double burden of excess adiposity and anaemia. DESIGN: Cross-sectional study, stratified two-stage cluster sample. BMI (=weight/height2) ≥25·0 kg/m2 defined overweight and BMI≥30·0 kg/m2 obesity. Anaemia was defined as Hb <120 g/l for women, <130 g/l for men. Gender inequalities vis-à-vis the within-subject coexistence of excess adiposity and anaemia were assessed by women v. men relative prevalence ratios (RPR). Their variation with sociodemographic characteristics used models including gender × covariate interactions. SETTING: Greater Tunis area in 2009-2010. SUBJECTS: Adults aged 20-49 years (women, n 1689; men, n 930). RESULTS: Gender inequalities in excess adiposity were high (e.g. overweight: women 64·9 % v. men 48·4 %; RPR=2·1; 95 % CI 1·6, 2·7) and much higher for anaemia (women 38·0 % v. men 7·2 %; RPR=8·2; 95 % CI 5·5, 12·4). They were striking for overweight and anaemia (women 24·1 % v. men 3·4 %; RPR=16·2; 95 % CI 10·3, 25·4). Gender inequalities in overweight adjusted for covariates increased with age but decreased with professional activity and household wealth score; gender inequality in anaemia or overweight and anaemia was more uniformly distributed. CONCLUSIONS: Women were much more at risk than men, from both over- and undernutrition perspectives. Both the underlying gender-related and sex-linked biological determinants of this remarkable double burden of malnutrition inequality must be addressed to promote gender equity in health.

A frequency questionnaire to estimate free-living physical activity among Tunisian preadolescent and adolescent children.

OBJECTIVE: To develop a child- and adolescent-appropriate physical activity frequency questionnaire (PAFQ) in Tunisia, North Africa. DESIGN: A PAFQ was developed from a physical activity (PA) inventory that comprised major activity components (at home, preparing meals, school time, transport, non-sport leisure, sports, prayer and sleeping time). Then, type and duration of each activity undertaken during the past week were estimated. Total energy expenditure (TEE) estimated by the PAFQ was compared with data derived from two criterion methods: heart-rate monitoring (HRM) and a 24 h PA recall (24h-R), both collected during a 3 d period including one weekday and two weekend days. SETTING: Two elementary schools and two high schools of the most developed and urbanized area, Greater Tunis. SUBJECTS: One hundred and forty-two volunteer children and adolescents aged 10-19 years. RESULTS: The PAFQ strongly was correlated with both HRM (r = 0·70; 95% CI 0·62, 0·76) and 24h-R (r = 0·81; 95% CI 0·77, 0·84). It featured acceptable agreement with both criterion measures, slightly underestimating TEE compared with 24h-R (-2·8%, mean of individual differences -272·7 kJ/d; 95% CI -490·6, -57·4 kJ/d) and moderately overestimating it compared with HRM (+11·3%, mean of individual differences +1106·2 kJ/d; 95% CI 845·8, 1366·6 kJ/d). Reliability ranged from moderate to good (weighted kappa coefficients from 0·47 to 0·78 and intra-class correlation coefficients between 0·79 and 0·86 for energy expenditure by PA categories), indicating strong agreement between the two assessments. CONCLUSIONS: This PAFQ could be useful in the description and surveillance of PA patterns or for the evaluation of population-based interventions directed at promoting PA in Tunisian children and adolescents.