Weekly expenditure on food and non-alcoholic beverages in Australia increased by an average of 16 % between 2009 and 2016(1). Contributing to this increase is the growing popularity of home-delivered meals, a trend that is also reflected in global food purchase patterns(Reference Maimaiti, Zhao and Jia2–5). Online meal delivery services are internet or application-based services through which consumers can order and receive ready-to-eat meals(Reference Ray, Dhir and Bala6). They include a wide range of delivery services, such as those provided by local restaurants and Uber Eats. Market research conducted by Roy Morgan has found that approximately one in five Australians (aged 14+ years) order a meal online at least once every 3 months, and that this prevalence rate has nearly doubled in the last 2 years(4,7) .
Individuals who more frequently consume meals that are cooked outside the home are more likely to have poorer diets and to gain weight(Reference Drewnowski8–Reference Tiwari, Aggarwal and Tang11). The trend towards online ordering is therefore problematic if the foods provided are typically unhealthy. A recent study conducted in three cities (Chicago, Amsterdam and Melbourne) suggested that this is indeed the case, with the most commonly advertised meals for home delivery being unhealthy options (e.g. burgers and pizzas)(Reference Poelman, Thornton and Zenk12).
Most research to date on the use of online meal ordering services has focused on (i) consumers’ attitudes towards the services and (ii) the capacity of restaurants to utilise these platforms(Reference Cho, Bonn and Li13–Reference Yeo, Goh and Rezaei17). In addition, some studies have assessed the demographic factors associated with use of online food ordering services, with this work finding that young adults, people residing in cities and people with previous experience ordering goods and services online being the most frequent users(Reference Maimaiti, Zhao and Jia2,7) . There appears to be a lack of information on other user characteristics to identify groups of consumers who are most likely to adopt this form of food purchasing. Examples of such characteristics include fruit, vegetable, fast-food and soft drink consumption; alcohol and tobacco use; engagement in physical activity; and sleep adequacy. Such information would provide insight into the extent of use of online meal delivery services and identifying groups of consumers that may most benefit from interventions designed to (i) encourage healthier food choices when ordering home-delivered meals and/or (ii) provide targeted information to service users about quick, tasty and healthy meals that can be prepared at home, thereby reducing the attractiveness of externally supplied options. In addition, more detailed profiling of these groups would facilitate tracking of the extent to which providers of unhealthy foods are targeting specific types of consumers. To address this research gap, the present study assessed a broad range of demographic and lifestyle factors to identify those who are most likely to engage in online meal ordering and thus could be the focus of future intervention efforts.
Methods
Study participants
Australian adults aged 18+ years (n 2010) completed an online survey disseminated via a web panel provider (Pureprofile). Data collection occurred from 23 July to 15 August 2018. The sample was nationally representative in terms of gender, age, socio-economic status and location (metropolitan v. regional).
Measures
Socio-demographics
Age, gender, education level (1 – ≤Year 11 to 7 – ≥postgraduate diploma or above), and annual gross household income (1 – ≤$9999/year to 5 – ≥$125 000/year) were assessed. In addition, respondents were asked to report whether they were the parent of a child younger than 18 years of age.
Physical health factors
Self-reported weight and height were used to calculate BMI (kg/m2). Self-rated health (1 – poor to 5 – excellent) and sleep adequacy (1 – rarely to 5 – daily) were also assessed.
Physical activity
Respondents reported the total amount of time spent doing moderate and vigorous physical activity in the last week. Those who engaged in ≥150 min/week of moderate and vigorous physical activity were classified as meeting the physical activity guideline(18).
Smoking
Respondents were asked to report their smoking status. Those who did not smoke regularly (regular smoking defined as consuming at least one cigarette/d or 7/week) were classified as meeting the relevant health recommendation.
Alcohol intake
Average alcohol intake was calculated by combining the items assessing frequency of alcohol consumption (1 – less than once a month to 7 – every day) and the number of standard drinks consumed on a typical drinking occasion (1 – half a drink to 11 – 20+ drinks)(19). Those consuming an average of ≤2 standard drinks of alcohol per d were classified as meeting the long-term risk guideline(20). Respondents consuming ≤4 standard drinks on a single occasion in the past 12 months were classified as meeting the short-term risk guideline(20).
Food preparation and consumption behaviours
Respondents’ usual intake of fruit and vegetables was assessed. Those consuming ≥5 servings of vegetables/d and ≥2 servings of fruit/d were classified as meeting the respective intake guidelines(21). Respondents reported the extent to which they enjoy cooking, consider themselves to be a good cook and feel confident to cook a nutritious meal (5-point scale from 1 – strongly disagree to 5 – strongly agree). Further items assessed the number of days respondents consumed sugary drinks (‘During the past 7 d, on how many days did you drink a can, bottle, or glass of a sugary drink such as soft drinks (like Coke/lemonade), energy drinks (like Red Bull), fruit drinks, sports drinks (like Gatorade), and cordial?’) and ate at a fast-food restaurant in the week prior to the survey (‘During the past 7 days, on how many days did you eat at a fast food outlet?’). Any special dietary requirements due to health or weight loss concerns were also assessed.
Online meal delivery services
Respondents were asked how often they had ordered meals online in the previous month. For analysis purposes, responses were dichotomised (0 – did not order meals online in the month prior to the survey v. 1 – ordered a meal online at least once in the month prior to the survey).
Statistical analyses
A series of univariate logistic regression analyses was conducted to assess the relationship between each predictor variable and the use of online meal ordering services at least once in the previous month. Variables found to be significant (P < 0·05) were included in a multivariate logistic regression model. Analyses were conducted using Stata version 16.1(22).
Results
Just over one in four (28 %) respondents ordered a meal via an online service in the month prior to the survey, with 10 % ordering on average once a week. The mean frequency of ordering meals online across the entire sample was once/month, increasing to once/week among those who had ordered in the previous month (n 556). Descriptive characteristics of the study population and measures are presented in Table 1.
Table 1 Descriptive characteristics of the study sample and factors measured (n 2010)
MVPA, moderate and vigorous physical activity.
Univariate logistic analyses revealed that the following factors were significantly associated with the use of online meal delivery services: age, gender, income, education, BMI, being a parent, meeting fruit intake guideline, meeting the moderate and vigorous physical activity guideline, frequency of fast-food intake, frequency of sugary drinks intake, meeting the short-term alcohol intake guideline, self-rated health, sleep adequacy, enjoying preparing food and cooking confidence. The subsequent multivariate logistic regression model showed that younger respondents, those with a higher BMI, and those with higher levels of education and income were more likely to use these services at least once a month (see Table 2). Additionally, the likelihood of using these services at least monthly was higher among those who reported consuming fast-food and sugary drinks more frequently.
Table 2 Factors associated with the use of online or app-based meal delivery services
MVPA, moderate and vigorous physical activity.
‘–’ indicates a non-significant association for univariate logistic regression analyses.
Discussion
Just over a quarter of respondents used online meal ordering services at least once in the previous month, and one in ten used them weekly. In line with prior research, younger and overweight or obese respondents, and those with higher education and income levels were found to be more likely to use these services(Reference Maimaiti, Zhao and Jia2,4,7,Reference Stephens, Miller and Militello23,Reference Keeble, Adams and Sacks24) . Consistent with previous research showing that home-delivered meals tend to be unhealthy(Reference Poelman, Thornton and Zenk12), the present study found that the use of home delivery services was greater among those consuming soft drinks and eating at fast-food restaurants more often.
In combination with previous work, the results suggest that the use of online meal ordering services may be becoming a mainstream activity in Australia and globally(4,7) , a trend that appears to have since intensified with the COVID-19 pandemic(25). While public health practitioners and policy-makers have introduced some strategies to help control the proliferation of fast-food establishments and assist the public to make healthier food choices in Australia and other countries(Reference Keeble, Burgoine and White26–Reference Morley, Scully and Martin28), online services present new challenges. For example, customers have access to product labelling in supermarkets and energy labelling on menus in some fast-food outlets, but in most Australian states there is no requirement on food delivery services to provide nutrition information. This has also been reported in UK(Reference d’Angelo, Gloinson and Draper29). It will become increasing important to address this deficit as more people choose to consume home-delivered meals.
Recent research provides some indication of how online purchase environments can be altered to encourage healthier selections. The strategies identified in online grocery shopping contexts include explicitly drawing attention to healthier choices, increasing the ratio of available healthier options relative to unhealthier options and providing filters that enable shoppers to refine their searches according to specific nutrition-related criteria(Reference Marty, Cook and Piernas30,Reference Olzenak, French and Sherwood31) . It remains to be seen whether such strategies would be effective in the online meal ordering context, and particularly among those population segments identified in this and previous research as being most likely to engage in this form of food purchase.
Given the difficulties associated with requiring food providers to modify their product ranges and labelling practices, advocacy can play an important role in bringing to bear the pressure required to encourage positive changes in marketing practices. This encouragement could take the form of ‘name and congratulate/shame’ reporting that publicises the providers with the best and worst nutritional profiles for their meal offerings(Reference Howes, Shahid and Jones32). Providing incentives in the form of recognition or endorsement has been found to improve the range of healthier ready-to-eat meals sold by food outlets(Reference Hillier-Brown, Summerbell and Moore33). In some instances, mandatory restrictions may be feasible due to the vulnerability of specific population segments. The findings of the present study are consistent with those from research conducted elsewhere indicating that online ordering is especially prevalent among younger people(Reference Maimaiti, Zhao and Jia2,Reference Stephens, Miller and Militello23,Reference Keeble, Adams and Sacks24) , making this group a priority segment for the development and implementation of appropriate interventions. For example, international evidence highlights the importance of introducing regulatory frameworks that limit the penetration of online ordering systems in settings where young people congregate, especially schools(Reference Stephens, Miller and Militello23).
Study strengths and limitations
The present study provides unique insights into which population subgroups are using online meal ordering services in a national sample of Australian adults. Although the data were collected prior to the COVID-19 pandemic, the results are timely given the substantially increased reliance on food deliveries during this period and the potential for this trend to continue into the future.
The primary limitation of this study was its cross-sectional design, which prevents determination of causal relationships between the assessed food consumption behaviours and the use of online meal ordering services. A second limitation was the use of a web panel for recruitment, which may have resulted in a sample that was skewed on relevant non-assessed variables (e.g. technology readiness). Future research may use different recruitment methods and a prospective design to provide additional information on factors predicting the use of online meal ordering services. Third, the data were self-reported and therefore potentially subject to social desirability bias. The final limitation was that this study was conducted in one country, Australia, and the results may not apply to the situations in other countries. Additional research could be undertaken in other national contexts to assess the extent of potential generalisability.
Conclusion
Online meal ordering is increasing in popularity in Australia, which is likely to have implications for diet quality at the population level. The present study found use of online ordering services to be greatest among younger people, those with higher levels of education and income, those of higher BMI, and those consuming more fast food and sugary drinks. Overall, the results highlight the need for evidence-based public health nutrition interventions and policies to be introduced in this space to ensure the increasing number of people using online meal ordering services have access to the nutrition information they need to make informed choices.
Acknowledgements
Acknowledgements: Not applicable. Financial support: This study was funded by Cancer Council Victoria and Cancer Council WA. Conflicts of interest: The authors declare that they have no conflicts of interest. Authorship: LD performed the data analysis. S.P., L.D., A.M. and A.B. conceptualised the study. L.D., E.H. and S.P. prepared the manuscript. A.M. and A.B. critically revised the paper for important intellectual content. All authors had access to all of the data in the study and can take responsibility for the integrity of the data and the accuracy of the data analysis. All authors have read and approved the final manuscript. Ethics of human subject participation: The study was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving research study participants were approved by the Curtin University Human Research Ethics Committee. All participants provided informed consent.
