The study examined the association between depressive symptoms and iron status, anaemia, body weight and pubertal status among Mexican adolescent girls.

In this cross-sectional study, depressive symptoms were assessed by the 6-item Kutcher Adolescent Depression Scale, and latent class analysis (LCA) was used to identify and characterise groups of girls based on depressive symptoms. Iron status and inflammation were assessed using ferritin and soluble transferrin receptor, C-reactive protein and alpha-1-acid glycoprotein, respectively. Multiple logistic and linear regressions were applied to model class membership as a function of iron status, anaemia, body weight and pubertal status.

We collected data from 408 girls aged 12–20 years.

Public schools in northern Mexico.

LCA yielded three classes of depressive symptoms: 44·4 % of the adolescents were ‘unlikely to be depressed’, 41·5 % were ‘likely to be depressed’ and 14·1 % were ‘highly likely to be depressed’. Our analyses demonstrated that iron-deficient girls had greater odds of being ‘likely depressed’ (OR 2·01, 95 % CI 1·01, 3·00) or ‘highly likely depressed’ (OR 2·80, 95 % CI 1·76, 3·84). Linear regression analyses revealed that lower Hb concentrations and higher body weight increased the probability of being ‘likely depressed’. There was no evidence that depressive symptoms were associated with age at menarche and years since menstruation.

This study shows that iron-deficient adolescent girls are more likely to suffer from depressive symptoms and that lower concentrations of Hb and higher body weight increased the probability of experiencing depressive symptoms.

Food advertising is an important determinant of unhealthy eating. However, analysing a large number of advertisements (ads) to distinguish between food and non-food content is a challenging task. This study aims to develop a machine learning-based method to automatically identify and classify food and non-food ad videos.

Methodological study to develop an algorithm model that prioritises both accuracy and efficiency in monitoring and classifying advertising videos.

From a collection of Brazilian television (TV) ads data, we created a database and split it into three sub-databases (i.e. training, validation and test) by extracting frames from ads. Subsequently, the training database was classified using the EfficientNet neural network. The best models and data-balancing strategies were investigated using the validation database. Finally, the test database was used to apply the best model and strategy, and results were verified with field experts.

The study used 2124 recorded Brazilian TV programming hours from 2018 to 2020. It included 703 food ads and over 20 000 non-food ads, following the protocol developed by the INFORMAS network for monitoring food marketing on TV.

The results showed that the EfficientNet neural network associated with the balanced batches strategy achieved an overall accuracy of 90·5 % on the test database, which represents a reduction of 99·9 % of the time spent on identifying and classifying ads.

The method studied represents a promising approach for differentiating food and non-food-related video within monitoring food marketing, which has significant practical implications for researchers, public health policymakers, and regulatory bodies.

To investigate clustering of risk behaviours in adolescents with excess weight.

Cross-sectional analysis of baseline data from the PRALIMAP-INÈS trial. Information on food frequency consumption (fruit, vegetables, sugary products and beverages), physical activity, sedentary behaviour (week and weekend days), smoking and alcohol consumption (current frequency and intoxication episodes) and socio-demographic data was collected using self-reported questionnaires. Behavioural risk factors were entered as categorical variables in a two-step clustering procedure: multiple correspondence analysis followed by hierarchical clustering. Associations between cluster membership and socio-demographic variables were investigated using multivariable multinomial logistic regression.

French PRALIMAP-INÈS trial.

Adolescents with excess weight.

A total of 1391 participants (13–18 years old, 58·2 % female) were included in the analysis, which resulted in the identification of four groups of participants, including, respectively, 543 (39·0 %), 373 (26·8 %), 246 (17·7 %) and 229 (16·5 %) participants. Clusters 1 and 4 showed associations of rather healthy behaviours (high physical activity and low consumption of sugary products; high consumption of fruit and vegetables, respectively), while clusters 2 and 3 showed associations of rather unhealthy behaviours (high sedentary behaviour and low consumption of fruit and vegetables; smoking and alcohol consumption, respectively). Both social status and family structure were associated with cluster membership.

Risk behaviour patterns in adolescents with excess weight were clustered in both healthier and less healthy ways, with a complex interplay with socio-demographic factors.

The objective of this study was to determine the association between the consumption of dairy foods with urinary iodine concentration (UIC) and iodine deficiency risk in a nationally representative sample of the US population.

24-hour dietary recall data and laboratory data for UIC (μg/l) from subjects 2+ years old US population participating in National Health and Nutrition Examination Surveys 2001–2018 were used (n 26 838) for analyses after adjusting for demographic covariates. Significant associations were assessed at P < 0·05.

Mean intakes of total dairy were 2·21, 2·17 and 1·70 cup equivalents (cup eq) among those 2–8, 9–18 and 19+ years, respectively. Of the dairy components, intake of milk was highest followed by cheese and yogurt for all age groups. Total dairy intakes were positively associated with UIC among those 2–8 years (β = 29·9 ± 9·9 μg/l urine/cup eq dairy) and 9–18 years (β = 26·0 ± 4·8 μg/l urine/cup eq dairy) but not associated among those 19+ years. Total dairy intakes were associated with lowered risks (30 %, 21 % and 20 % for among 2–8, 9–18 and 19+ years, respectively) of being classified as iodine insufficient (UIC < 100 μg/l) or lowered risk (47 %, 30 % and 26 % among 2–8, 9–18 and 19+ years, respectively) of being classified as iodine severely deficient (UIC < 20 μg/l).

The results indicate that dairy foods are beneficially associated with UIC and lowered iodine deficiency risk.

To examine energy drink consumption among adolescents in the UK and associations with deprivation and dietary inequalities.

Quantitative dietary and demographic data from the National Diet and Nutrition Survey (NDNS) repeated cross-sectional survey were analysed using logistic regression models. Qualitative data from semi-structured interviews were analysed using inductive thematic analysis.

UK.

Quantitative data: nationally representative sample of 2587 adolescents aged 11–18 years. Qualitative data: 20 parents, 9 teachers and 28 adolescents from Hampshire, UK.

NDNS data showed adolescents’ consumption of energy drinks was associated with poorer dietary quality (OR 0·46 per sd; 95 % CI (0·37, 0·58); P < 0·001). Adolescents from more deprived areas and lower income households were more likely to consume energy drinks than those in more affluent areas and households (OR 1·40; 95 % CI (1·16, 1·69); P < 0·001; OR 0·98 per £1000; 95 % CI (0·96, 0·99); P < 0·001, respectively). Between 2008 and 2016, energy drink consumption among adolescents living in the most deprived areas increased, but decreased among those living in the most affluent neighbourhoods (P = 0·04). Qualitative data identified three themes. First, many adolescents drink energy drinks because of their friends and because the unbranded drinks are cheap. Second, energy drink consumption clusters with other unhealthy eating behaviours and adolescents do not know why energy drinks are unhealthy. Third, adolescents believe voluntary bans in retail outlets and schools do not work.

This study supports the introduction of age-dependent legal restrictions on the sale of energy drinks which may help curb existing socio-economic disparities in adolescents’ energy drink intake.

To assess the alignment of YouTube® videos providing dietary recommendations for gout with evidence-based guidelines targeted at the United Kingdom (UK) population and to establish their quality.

A content analysis of YouTube® videos providing dietary recommendations for gout was undertaken. Videos were categorised by video source. Each video’s dietary recommendations for gout were compared with three evidence-based guidelines for gout, producing a compliance score. Presence of non-guideline advice was assessed. Understandability and actionability were evaluated using the Patient Education Material Assessment Tool for Audio-Visual Materials. Reliability was assessed using an adapted-DISCERN tool and educational quality using the Global Quality Score Five-Point Scale. Differences between video source and continuous variables were assessed using one-way Kruskal–Wallis H tests. For categorical variables, associations were investigated using Fisher–Freeman–Halton tests.

Online, May–June 2020.

One-hundred thirty-one videos.

Alignment of videos with evidence-based guidelines was poor (median compliance score 27 % (interquartile range 17–37 %)). Additionally, 57 % of videos contained non-guideline advice. The health professional source group had the fewest videos containing non-guideline advice, but this was only significantly lower than the naturopath group (31 % v. 81 %, P = 0·009). Almost 70 % of videos were considered poorly actionable and 50 % poorly understandable. Most videos were rated poor for reliability (79 %) and poor to generally poor for educational quality (49 %).

YouTube® videos providing dietary recommendations for gout frequently fail to conform to evidence-based guidelines, and their educational quality, reliability, understandability and actionability are often poor. More high-quality, comprehensive, evidence-based YouTube® videos are required for UK gout patients.

We evaluated associations between food insecurity (FI) and the quality and quantity of sleep in adults (≥18 years).

The current study represented a systematic review and meta-analysis of observational studies.

Databases of PubMed, Scopus, Embase and Web of Science were searched from inception until 6 June 2022. Meta-analyses were conducted using random-effects models, and effect sizes were reported as OR and 95 % CI.

Data from ten eligible observational studies, including 83 764 participants, were included.

FI was associated with an increased risk of poor sleep quality (OR = 1·45; 95 % CI (1·24, 1·70), I2 = 95, P < 0·001, n 7). Besides, subgroup analysis showed increased risk of poor sleep quality corresponding to the severity of FI across mild (OR = 1·31; 95 % CI (1·16, 1·48), I2 = 0 %, P < 0·001, n 5), moderate (OR = 1·49; 95 % CI (1·32, 1·68), I2 = 0 %, P < 0·001, n 5) and severe (OR = 1·89; 95 % CI (1·63, 2·20), I2 = 0 %, P < 0·001, n 5) levels. Similarly, subgroup analysis by sleep problems showed that FI was associated with an increased the risk of trouble falling asleep (OR = 1·39; 95 % CI (1·05, 1·83), I2 = 91 %, P = 0·002, n 3) and trouble staying asleep (OR = 1·91; 95 % CI (1·37, 2·67), I2 = 89 %, P < 0·001, n 3). Moreover, FI was associated with the odds of shorter (OR = 1·14; 95 % CI (1·07, 1·21), I2 = 0 %, P < 0·001, n 4) and longer sleep duration (OR = 1·14; 95 % CI (1·03, 1·26), I2 = 0 %, P = 0·010, n 4).

Collective evidence supports that FI is associated with poor sleep quality and quantity in adults. Preventative and management strategies that address FI may provide health benefits beyond improving nutritional status per se.

This study examines the individual and combined association of BMI and waist-to-hip ratio (WHR) with CVD risk using genetic scores of the obesity measurements as proxies.

A 2 × 2 factorial analysis approach was applied, with participants divided into four groups of lifetime exposure to low BMI and WHR, high BMI, high WHR, and high BMI and WHR based on weighted genetic risk scores. The difference in CVD risk across groups was evaluated using multivariable logistic regression.

Cohort study.

A total of 408 003 participants were included from the prospective observational UK Biobank study.

A total of 58 429 CVD events were recorded. Compared to the low BMI and WHR genetic scores group, higher BMI or higher WHR genetic scores were associated with an increase in CVD risk (high WHR: OR, 1·07; 95 % CI (1·04, 1·10)); high BMI: OR, 1·12; 95 % CI (1·09, 1·16). A weak additive effect on CVD risk was found between BMI and WHR (high BMI and WHR: OR, 1·16; 95 % CI (1·12, 1·19)). Subgroup analysis showed similar patterns between different sex, age (<65, ≥65 years old), smoking status, Townsend deprivation index, fasting glucose level and medication uses, but lower systolic blood pressure was associated with higher CVD risk in obese participants.

High BMI and WHR were associated with increased CVD risk, and their effects are weakly additive. Even though there were overlapping of effect, both BMI and WHR are important in assessing the CVD risk in the general population.

To investigate the association between folate levels and the risk of gestational diabetes mellitus (GDM) risk during the whole pregnancy.

In this retrospective cohort study of pregnant women, serum folate levels were measured before 24 gestational weeks (GW). GDM was diagnosed between 24th and 28th GW based on the criteria of the International Association of Diabetes and Pregnancy Study Groups. General linear models were performed to examine the association of serum folate with plasma glucose (i.e. linear regressions) and risk of GDM (i.e. log-binomial regressions) after controlling for confounders. Restricted cubic spline regression was conducted to test the dosage–response relationship between serum folate and the risk of GDM.

A sigle, urban hospital in Shanghai, China.

A total of 42 478 women who received antenatal care from April 2013 to March 2017 were included.

Consistent positive associations were observed between serum folate and plasma glucose levels (fasting, 1-h, 2-h). The adjusted relative risks (RR) and 95 % CI of GDM across serum folate quartiles were 1·00 (reference), 1·15 (95 % CI (1·04, 1·26)), 1·40 (95 % CI (1·27, 1·54)) and 1·54 (95 % CI (1·40, 1·69)), respectively (P-for-trend < 0·001). The positive association between serum folate and GDM remained when stratified by vitamin B12 (adequate v. deficient groups) and the GW of serum folate measurement (≤13 GW v. >13 GWs)

The findings of this study may provide important evidence for the public health and clinical guidelines of pregnancy folate supplementation in terms of GDM prevention.

To compare ultra-processing markers and nutrient composition in plant-based meat products (PBMP) with equivalent meat-based products (MBP).

A total of 282 PBMP and 149 MBP within 18 product categories were assessed. Based on the NOVA classification, 33 ultra-processing markers were identified and six ultra-processing bullet categories were defined, that is flavour, flavour enhancer, sweetener, colour, other cosmetic additives and non-culinary ingredients. The ingredient lists were analysed concerning these ultra-processing markers and ultra-processing bullet categories, as well as nutrient composition, for all PBMP and MBP. Differences between PBMP and MBP were assessed using chi-square and Mann-Whitney U tests, respectively.

Cross-sectional analysis.

282 PBMP and 149 MBP.

The percentage of ultra-processed food (UPF) items was significantly higher in PBMP (88 %) as compared to MBP (52 %) (P < 0·0001). The proportion of UPF items was numerically higher in 15 out of 18 product categories with differences in six categories reaching statistical significance (P < 0·05). Flavour, flavour enhancer, colour, other cosmetic additives and non-culinary ingredients were significantly more prevalent in PBMP as compared to MBP (P < 0·0001). Concerning nutrient composition, median energy, total fat, saturated fat and protein content were significantly lower, whereas the amounts of carbohydrate, sugar, fibre and salt were significantly higher in PBMP (P < 0·05).

Ultra-processing markers are significantly more prevalent in PBMP as compared to MBP. Since UPF intake has been convincingly linked to metabolic and CVD, substituting MBP with PBMP might have negative net health effects.

To document perfluoroalkyl acids (PFAA) and bisphenol-A (BPA) exposure in four First Nation communities in northern Quebec compared with the Canadian Health Measures Survey (CHMS Cycle 5 2016–2017) and examine the associations between dietary consumption and chemical exposure.

We used cross-sectional data from the JES-YEH! project conducted in collaboration with four First Nation communities in 2015. A FFQ collected information on diet, and PFAA and BPA were measured in biological samples. We used generalised linear models to test the associations between food intake and chemical biomarkers.

Northern Quebec.

Youth aged 3–19 years (n 198).

Mean perfluorononanoic acid (PFNA) levels were significantly higher in JES-YEH! than CHMS, and BPA levels were higher among those aged 12–19 years compared with CHMS. Dairy products were associated with PFNA among Anishinabe and Innu participants (geometric mean ratio 95 % CI: 1·53 (95 % CI 1·03, 2·29) and 1·52 (95 % CI 1·05, 2·20), respectively). PFNA was also associated with ultra-processed foods (1·57 (95 % CI 1·07, 2·31)) among Anishinabe, and with wild fish and berries (1·44 (95 % CI 1·07, 1·94); 1·75 (95 % CI 1·30, 2·36)) among Innu. BPA was associated with cheese (1·72 (95 % CI 1·19, 2·50)) and milk (1·53 (95 % CI 1·02, 2·29)) among Anishinabe, and with desserts (1·71 (95 % CI 1·07, 2·74)), processed meats (1·55 (95 % CI 1·00, 2·38)), wild fish (1·64 (95 % CI 1·07, 2·49)) and wild berries (2·06 (95 % CI 1·37, 3·10)) among Innu.

These results highlight the importance of better documenting food-processing and packaging methods, particularly for dairy products, and their contribution to endocrine disruptors exposures as well as to promote minimally processed and unpackaged foods to provide healthier food environments for youth in Indigenous communities and beyond.

Night shift workers are at a 20 to 40 % increased risk of metabolic diseases, which may be associated with their disrupted eating patterns. This qualitative study explores factors that influence night shift workers’ eating habits, within and outside of the workplace, to identify target areas for health promotion strategies.

Eligible participants resided in Australia, working at least three overnight shifts per month.

The photovoice method was used, whereby participants were asked to take photos that represent their typical eating habits. These photos were incorporated as prompts in a semi-structured interview, which explored factors influencing eating habits on night shifts and days-off and perceptions and enablers to healthy eating.

Ten participants completed the study. Thematic analysis generated four main themes, which were mapped onto the Social Ecological Model (SE Model). Aligned with the SE Model, our results show night shift workers’ eating habits are influenced by intrapersonal, interpersonal and (work) organisational levels. Participants reported that at work, appropriate food preparation facilities are required to enable healthy food choices. Poor shift work rostering leads to prolonged fatigue on days-off, limiting their ability and motivation to engage in healthy eating. Consequently, night shift workers seem to require additional supports from their social networks and enhanced food literacy skills, in order to adopt/ maintain healthy eating behaviours.

Night shift work creates individual and environmental barriers to healthy eating, which are present during and outside of work. Health promotion strategies for this population should include multiple approaches to address these barriers.

To quantify the dose–response relation between yogurt consumption and risk of mortality from all causes, CVD and cancer.

Systematic review and meta-analysis.

We conducted a comprehensive search of PubMed/Medline, ISI Web of Science and Scopus databases through August 2022 for cohort studies reporting the association of yogurt consumption with mortality from all causes, CVD and cancer. Summary relative risks (RR) and 95 % CI were calculated with a random-effects model.

Seventeen cohort studies (eighteen publications) of 896 871 participants with 75 791 deaths (14 623 from CVD and 20 554 from cancer).

High intake of yogurt compared with low intake was significantly associated with a lower risk of deaths from all causes (pooled RR 0·93; 95 % CI: 0·89, 0·98, I2 = 47·3 %, n 12 studies) and CVD (0·89; 95 % CI: 0·81, 0·98, I2 = 33·2 %, n 11), but not with cancer (0·96; 95 % CI: 0·89, 1·03, I2 = 26·5 %, n 12). Each additional serving of yogurt consumption per d was significantly associated with a reduced risk of all-cause (0·93; 95 % CI: 0·86, 0·99, I2 = 63·3 %, n 11) and CVD mortality (0·86; 95 % CI: 0·77, 0·99, I2 = 36·6 %, n 10). There was evidence of non-linearity between yogurt consumption and risk of all-cause and CVD mortality, and there was no further reduction in risk above 0·5 serving/d.

Summarising earlier cohort studies, we found an inverse association between yogurt consumption and risk of all-cause and CVD mortality; however, there was no significant association between yogurt consumption and risk of cancer mortality.

To evaluate the relationships of fibre intake with subsequent BMI sd-score, waist-to-height ratio and serum fasting glucose levels among school-age Japanese children.

This is a prospective study of school-age Japanese children. Participants were followed from 6–7 to 9–10 years of age (follow-up rate: 92·0 %). Fibre intake was assessed using a validated FFQ. Serum fasting glucose was measured by a hexokinase enzymatic method. Using a general linear model, the associations between dietary fibre intake at baseline and BMI sd-score, waist-to-height ratio, and serum levels of fasting glucose at follow-up were evaluated after considering potential confounding factors.

Public elementary schools in a city in Japan

A total of 2784 students.

The estimated means for fasting glucose at 9–10 years of age were 86·45, 85·68, 85·88 and 85·58 mg/dl in the lowest, second, third and highest quartile of fibre intake at 6–7 years of age, respectively (P = 0·033, trend P = 0·018). Higher fibre intake at 6–7 years of age was associated with lower waist-to-height ratio at 9–10 years of age (trend P = 0·023). The change in fibre intake was inversely associated with concurrent change of BMI sd-score (trend P = 0·044).

These results suggest that dietary fibre intake may be potentially effective to limit excess weight gain and lower glucose levels during childhood.

We investigated the relationship between socio-economic status and child undernutrition in West Africa (WA), and further examined the mediating role of dietary practices (measured as minimum dietary diversity (MDD), minimum meal frequency (MMF) and minimum acceptable diet (MAD)) and household environmental quality (HEQ) in the observed relationship.

Thirteen countries were included in the study. We leveraged the most recent Demographic and Health Surveys datasets ranging from 2010 to 2019. Poisson regression model with robust standard errors was used to estimate prevalence ratios and their corresponding 95 % CI. Structural equation modelling was used to conduct the mediation analysis.

West Africa.

132 448 under-five children born within 5 years preceding the survey were included.

Overall, 32·5 %, 8·2 %, 20·1 % and 71·7 % of WA children were stunted, wasted, underweight and anaemic, respectively. Prevalence of undernutrition decreased with increasing maternal education and household wealth (Trend P-values < 0·001). Secondary or higher maternal education and residence in rich households were associated with statistically significant decrease in the prevalence of stunting, wasting, underweight and anaemia among children in WA. MAD was found to mediate the association of low maternal education and poor household wealth with childhood stunting and underweight by 35·9 % to 44·5 %. MDD, MMF and HEQ did not mediate the observed relationship.

The study findings enables an evaluation and improvement of existing intervention strategies through a socio-economic lens to help address the high burden of child undernutrition in WA and other developing regions.

The goal of this study is to evaluate university students’ perceptions of tap water safety and water filter use and determine how these perceptions and behaviours affect water and sugar-sweetened beverage intake.

Cross-sectional; online survey conducted in Fall 2021.

A large, public Midwestern university in the USA.

Seven-hundred ninety-three university students.

Students who experienced food insecurity, were on a Pell grant, were first-generation college students or were racial/ethnic minorities were less likely to trust tap water safety. Tap water filtration behaviour also varied by age and race/ethnicity. Students who did not agree with the statement ‘my local tap water is safe to drink’ had lower odds of consuming ≥ 3 cups of total water per day (OR = 0·45, 95 % CI: 0·32, 0·62), lower odds of consuming tap water ≥ 3 times/d (OR = 0·46, 95 % CI: 0·34, 0·64), higher odds of drinking bottled water ≥ 1 time per day (OR = 1·80, 95 % CI: 1·22, 2·66) and higher odds of drinking SSB ≥ 1 time per day (OR = 1·47, 95 % CI: 1·01, 2·14) than those who agreed. Students who always or sometimes filtered their tap water had lower odds of consuming ≥ 3 cups of total water per day (OR = 0·59, 95 % CI: 0·39, 0·90) than students who never filtered their tap water.

Tap water perceptions and behaviours affect tap and bottled water and SSB intake among university students. Tap water perceptions and behaviours in this demographic provide important context for university programming promoting healthy beverage initiatives.

To evaluate the associations of ultra-processed food (UPF) consumption and obesity indicators among individuals with and without type 1 diabetes mellitus (T1DM) from the Coronary Artery Calcification in Type 1 Diabetes cohort study.

A secondary analysis. The consumption of UPF was assessed using the dietary data collected with the Harvard FFQ, and each food item was categorised according to the NOVA food processing classification. Height, weight and waist circumference were measured at baseline and after a mean of 14·6-year follow-up. Generalised estimating equations stratified by diabetes status were used to assess the associations between UPF intake and obesity indicators over 14 years of follow-up.

USA.

A total of 600 adults (256 T1DM and 344 non-diabetic controls) aged 39 ± 9·1 years at baseline and followed up for over 14 years were included.

Participants with T1DM consumed significantly more UPF than non-diabetic controls at baseline: 7·6 ± 3·8 v. 6·6 ± 3·4 servings per day of UPF, respectively (P < 0·01). Participants with T1DM and with the highest UPF intake had the highest weight (βQ4 v. Q1 = 3·07) and BMI (βQ4 v. Q1 = 1·02, all P < 0·05) compared with those with the lowest UPF intake. Similar positive associations were observed in non-diabetic controls.

Individuals with T1DM may consume more UPF than non-diabetic controls. Positive associations between UPF consumption and obesity indicators suggest that limiting UPF can be recommended for obesity prevention and management. Further research is needed to confirm these findings.

To examine associations between work and commute hours with food consumption and test whether neighbourhood type (20-minute neighbourhood (20MN)/non-20MN) moderate associations.

Cross-sectional analysis of the Places and Locations for Activity and Nutrition study (ProjectPLAN). Exposures were work hours (not working (0 h), working up to full-time (1–38 h/week), working overtime (> 38 h/week)), and among those employed, combined weekly work and commute hours (continuous). Outcomes were usual consumption of fruit, vegetables, takeaway food, snacks and soft drinks, and number of discretionary food types (takeaway, snacks and soft drinks) consumed weekly. Generalised linear models were fitted to examine associations between each exposure and outcome. The moderating role of neighbourhood type was examined through interaction terms between each exposure and neighbourhood type (20MN/non-20MN).

Melbourne and Adelaide, Australia, 2018–2019.

Adults ≥ 18 years old (n 769).

Although all confidence intervals contained the null, overall, patterns suggested non-workers and overtime workers have less healthy food behaviours than up-to-full-time workers. Among those employed, analysis of continuous work and commute hours data suggested longer work and commute hours were positively associated with takeaway consumption (OR = 1·014, 95 % CI 0·999, 1·030, P-value = 0·066). Patterns of better behaviours were observed across most outcomes for those in 20MN than non-20MN. However, differences in associations between work and commute hours with food consumption across neighbourhood type were negligible.

Longer work and commute hours may induce poorer food behaviours. There was weak evidence to suggest 20MN moderate associations between work and commute hours with food consumption, although behaviours appeared healthier for those in 20MN.