This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Formative Research, is properly cited. The complete bibliographic information, a link to the original publication on https://formative.jmir.org, as well as this copyright and license information must be included.
A poor diet contributes substantially to the development of noncommunicable diseases. In Singapore, it is recommended to consume at least 2 servings of fruits and vegetables daily to reduce the risk of developing noncommunicable diseases. However, the adherence rate among young adults is low. The COVID-19 pandemic has led to frequent users of mobile food delivery apps (MFDAs) adopting unhealthy eating habits, including high consumption of sugar-sweetened beverages, making it crucial to gain a deeper understanding of the underlying factors driving their use patterns.
We aimed to examine the use patterns of MFDAs among young adults during the COVID-19 pandemic; investigate the association between MFDA use and sociodemographic factors, dietary factors, and BMI; identify the underlying reasons for the observed use patterns of MFDAs among users; and compare the influences of MFDA use between frequent and infrequent users.
A sequential mixed methods design was used involving a web-based survey and in-depth interviews with a subset of respondents. Poisson regression and thematic analysis were used to analyze the quantitative and qualitative data, respectively.
The quantitative results revealed that 41.7% (150/360) of participants reported using MFDAs frequently, defined as at least once a week. Although not substantial, the study found that frequent users were less likely to consume 2 servings of vegetables per day and more likely to drink sugar-sweetened beverages. Nineteen individuals who had participated in the quantitative component were selected for and completed the interviews. Qualitative analysis identified 4 primary themes: deliberations about other sources of meals versus meals purchased via MFDAs, convenience is vital, preference for unhealthy meals ordered from MFDAs most of the time, and cost is king. Before making any purchase, MFDA users consider all these themes at the same time, with cost being the most important influential factor. A conceptual framework based on these themes was presented. Lack of culinary skills and COVID-19 restrictions were also found to influence frequent use.
This study suggests that interventions should focus on promoting healthy dietary patterns in young adults who frequently use MFDAs. Teaching cooking skills, especially among young male individuals, and time management skills could be useful to reduce reliance on MFDAs. This study highlights the need for public health policies that make healthy food options more affordable and accessible. Given the unintended changes in behavior during the pandemic, such as reduced physical activity, sedentary behavior, and altered eating patterns, it is essential to consider behavior change in interventions aimed at promoting healthy lifestyles among young adults who frequently use MFDAs. Further research is needed to evaluate the effectiveness of interventions during COVID-19 restrictions and assess the impact of the post–COVID-19
Poor dietary habits are a leading cause of death worldwide because of noncommunicable diseases (NCDs) [
The Health Promotion Board of Singapore, a government organization promoting healthy living, recommends consuming at least 2 servings of fruits and 2 servings of vegetables per day, with water as the preferred beverage [
Young adulthood is a crucial period for NCD prevention, and weight gain during this phase is a substantial risk factor for NCD development. A trend analysis indicated that the incidence of obesity among young adults was increasing, with a UK study showing that young adults aged 18 to 24 years had a 4-fold higher risk of transitioning from normal weight to overweight or obese in 10 years than those aged 65 to 74 years [
The COVID-19 pandemic has presented opportunities for increased adoption of mobile food delivery apps (MFDAs) as a safe and convenient food purchase method given the strict enforcement of social distancing measures [
Studies examining the use of MFDAs have shown varying prevalence rates and patterns across different countries. A study conducted in Western countries found that only 15% of participants reported using MFDAs at least once a week [
Other studies have also shown that MFDAs are often used to order unhealthy foods. A study conducted in Malaysia found that 77.6% of young adult users ordered unhealthy food, with 40.3% using MFDAs 1 to 3 times a month [
To the researchers’ knowledge, this is the first study that focuses on the use of MFDAs and its association with the healthfulness of users’ diets and explains the use patterns through qualitative interviews during the COVID-19 pandemic when restrictions were in place. This study’s research objectives were to (1) examine the use patterns of MFDAs among young adults during the COVID-19 pandemic; (2) investigate the association between MFDA use and sociodemographic factors, dietary factors, and BMI; (3) identify the underlying reasons for the observed use patterns of MFDAs among users; and (4) compare the influences of MFDA use between frequent and infrequent users.
This cross-sectional study adopted an explanatory sequential mixed methods research design to provide a comprehensive understanding of the research topic, which cannot be obtained through the conduct of a single quantitative or qualitative study [
The first 2 research objectives can be aptly addressed using a quantitative design, whereas the last 2 research objectives can be adequately answered through a qualitative design. In the
In the quantitative component, participants completed a questionnaire, and in the qualitative component, a subset of participants took part in semistructured in-depth interviews. To guarantee the quality and appropriate interpretation of the data, the researchers followed the CHERRIES (Checklist for Reporting Results of Internet E-Surveys) guidelines [
Participants from National University of Singapore (NUS) were recruited to complete the web-based questionnaire via convenience sampling. The eligibility criterion was students aged 18 to 35 years. The research team sent an email to each teaching department in NUS. The departments that agreed to participate forwarded this email to their students. This email contained a general link to a REDCap (Research Electronic Data Capture; Vanderbilt University) web page [
The selection criteria for the qualitative participants were the same as the eligibility criteria for the quantitative design. The study intentionally recruited individuals who used MFDAs more frequently compared with those who used them less often to gather insights and feedback on MFDA use patterns. In addition, these participants were selected via purposeful stratified sampling, reflecting the age, sex, and racial profile of young adults in Singapore [
Using a sample size calculator [
One of the researchers (XCT) created a sample questionnaire based on the literature [
Participants were asked questions about the frequency of MFDA use. A 10-point Likert scale ranging from “Never or rarely” to “6+ a day” was adapted from a diet screener [
Questions about other patterns included changes in MFDA use compared with before the COVID-19 pandemic, the period in which participants tended to purchase more often, and the types of cuisines commonly ordered.
Fruit and vegetable consumption was assessed using the following questions: (1) How many servings of fresh fruits do you eat on a typical day? and (2) How many servings of vegetables do you eat on a typical day? The choices were presented as an 8-point Likert scale ranging from “0” to “>6.” The Likert scale was adapted from the Singapore National Health Survey 2010 [
SSBs in this study were defined as any beverage that contained added sugar, including low-sugar or low-caloric drinks. SSB consumption was assessed using the following question: “How many servings of sugar-sweetened beverages do you consume on a typical day?” The choices were presented as an 8-point Likert scale ranging from “0” to “>6” adapted from the study by Robertson et al [
The self-reported height and weight of the participants were recorded. BMI was calculated as
A question about leisure-time physical activity level was adapted from the Singapore National Health Survey 2010 (“In the past 3 months, did you participate in any sports, exercise or walking during your leisure time?”) [
Participants’ demographics (including age, sex, race, marital status, number of children, working status, student status [undergraduate or graduate and nongraduating students], and study status [full-time or part-time study]) were collected.
The researchers formulated the initial questions of the interview guide to explore the research questions in an open-ended, inductive way (
The guide was vetted by public health nutritionists and reviewed by an academic whose teaching and research involve qualitative methods. XCT pilot-tested the guide on a university student, and further changes were made before it was sent for ethics approval.
What are the factors that drive frequent or infrequent use of MFDAs?
Probe: in your opinion, what are the factors that drive your personal use of MFDAs? Why do you think so?
How did your use patterns change during the COVID-19 pandemic?
Probe: how has your use of MFDAs changed from the pre–COVID-19 period to the current situation?
What attracts you to use certain MFDAs?
Probe: what are the factors that entice users to use certain food apps?
Why are there high use patterns observed for lunch and dinner?
Probe: as compared with eating out, what are the enticing factors for one to buy meals via MFDAs?
Why is food ordered via MFDAs mostly unhealthy?
Probe: do you feel that young adults purchase healthier or less healthy meals through the apps? What are the possible reasons?
Why do you think that the food ordered cannot meet the dietary requirements of fruits and vegetables?
Probe: do the foods purchased from the apps contain adequate fruits and vegetables? Why is this so?
How do users decide to purchase specialty drinks via MFDAs?
Probe: what might entice users to purchase sugar-sweetened beverages through the apps?
How has the frequent use of MFDAs affected your level of physical activity?
Probe: how has the physical activity level of MFDA users changed because of the convenience of food delivery?
How will your use patterns change once you start working?
How will your use patterns change once you are married and have children?
This study was approved by NUS SSHSPH Department Ethics Review Committee (SSHSPH-056). All participants were required to provide written informed consent to take part. The data collected were coded to keep the identities of the participants confidential.
Responses to the questionnaires were collected from January 14, 2021, to April 30, 2021, whereas qualitative interviews were conducted from March 18, 2021, to May 21, 2021, during a less restrictive phase of Singapore’s COVID-19 measures.
Before the start of the questionnaire, potential participants were informed that taking part in this project was completely voluntary and that their responses would contribute to the understanding of a new area relevant to promoting healthy eating among young adults. Potential participants were also given a guarantee of anonymity and were informed that only XCT and his research supervisors, FM-R and NAP, would have access to the raw data. The expected time to complete the questionnaire was 15 minutes. The end of the survey asked for their contact details if they were interested in participating in the qualitative interviews in the future. Once the participants had submitted their responses, they were unable to modify their answers. There were no cookies saved on the electronic devices of the participants, and the REDCap servers did not store any IP addresses. Participants were not given any reimbursements for completion of the questionnaire.
XCT conducted the face-to-face interviews via Zoom (Zoom Video Communications) [
All quantitative data were analyzed using Stata (version 17; StataCorp) [
All the interviews were transcribed verbatim. Qualitative data were analyzed using NVivo (QSR International) for Windows [
Of the 507 entries recorded in REDCap, 360 (71%) responses were retrieved for analysis after removal of duplicates (n=138, 27.2%) and incomplete data (n=9, 1.8%; see
Study flow diagram of how the participants were selected. REDCap: Research Electronic Data Capture.
Participants’ demographic profile.
Characteristic | Survey participants (n=360) | Interview participants (n=19) | |||
|
23 (21-25) | 25 (23.0-27.5) | |||
|
18-23, n (%) | 233 (64.7) | 6 (31.6) | ||
|
24-29, n (%) | 94 (26.1) | 10 (52.6) | ||
|
30-35, n (%) | 33 (9.2) | 3 (15.8) | ||
Female participants, n (%) | 249 (69.2) | 10 (52.6) | |||
Chinese participants, n (%) | 308 (85.6) | 13 (68.4) | |||
Not working, n (%) | 276 (76.7) | 11 (57.9) | |||
Single, n (%) | 337 (93.6) | 18 (94.7) | |||
No children, n (%) | 352 (97.8) | 18 (94.7) | |||
Undergraduates, n (%) | 248 (68.9) | 7 (36.8) | |||
Full-time study, n (%) | 314 (87.2) | 15 (78.9) |
Mobile food delivery app use patterns (n=360).
|
Participants, n (%) | ||
|
|||
|
|
210 (58.3) | |
|
|
Never or rarely | 53 (14.7) |
|
|
Once a month | 63 (17.5) |
|
|
2-3 times a month | 94 (26.1) |
|
Once a week | 70 (19.4) | |
|
|
80 (22.2) | |
|
|
2-3 times a week | 48 (13.3) |
|
|
4-6 times a week | 20 (5.6) |
|
|
Once a day | 6 (1.7) |
|
|
2-3 times a day | 4 (1.1) |
|
|
4-5 times a day | 2 (0.6) |
|
|||
|
No | 57 (18.6) | |
|
Yes | 250 (81.4) | |
|
|||
|
Breakfast | 26 (4.5) | |
|
Lunch | 224 (38.4) | |
|
Tea break | 44 (7.5) | |
|
Dinner | 220 (37.7) | |
|
Night snacks | 70 (12) | |
|
|||
|
Beverages | 109 (12.9) | |
|
Salads | 15 (1.8) | |
|
Desserts | 55 (6.5) | |
|
Asian-based | 249 (29.5) | |
|
Fast food | 225 (26.7) | |
|
Western-based | 182 (21.6) | |
|
Convenience food (eg, microwavable food and instant noodles) | 8 (0.9) |
aMissing values because of branching of question.
A higher percentage of female participants used MFDAs less frequently compared with their male counterparts (150/249, 60.2% vs 60/111, 54.1%), whereas a higher percentage of graduates and nongraduating students were more likely to use MFDAs frequently compared with undergraduates (once a week: 28/112, 25% vs 42/248, 16.9%; at least 2 times per week: 28/112, 25% vs 52/248, 21%). However, none of these differences were statistically significant (
Sociodemographic variables associated with frequency of mobile food delivery app (MFDA) use (n=360).
|
Frequency of MFDA use, n (%) | Chi-square ( |
||||||||
|
Less than once a week | Once a week | At least 2 times per week |
|
|
|||||
|
4.0 (4) | .41 | ||||||||
|
18-23 | 142 (60.9) | 41 (17.6) | 50 (21.5) |
|
|
||||
|
24-29 | 47 (50) | 23 (24.5) | 24 (25.5) |
|
|
||||
|
30-35 | 21 (63.6) | 6 (18.2) | 6 (18.2) |
|
|
||||
|
3.0 (2) | .22 | ||||||||
|
Nonfemalea | 60 (54.1) | 20 (18) | 31 (27.9) |
|
|
||||
|
Female | 150 (60.2) | 50 (20.1) | 49 (19.7) |
|
|
||||
|
5.0 (2) | .08 | ||||||||
|
Non-Chinese | 29 (55.8) | 6 (11.5) | 17 (32.7) |
|
|
||||
|
Chinese | 181 (58.8) | 64 (20.8) | 63 (20.5) |
|
|
||||
|
0.8 (2) | .69b | ||||||||
|
Married | 13 (56.5) | 6 (26.1) | 4 (17.4) |
|
|
||||
|
Single | 197 (58.5) | 64 (19) | 76 (22.6) |
|
|
||||
|
2.0 (2) | .30b | ||||||||
|
≥1 | 3 (37.5) | 3 (37.5) | 2 (25) |
|
|
||||
|
0 | 207 (58.8) | 67 (19) | 78 (22.2) |
|
|
||||
|
1.6 (2) | .45 | ||||||||
|
Yes | 48 (57.1) | 20 (23.8) | 16 (19.1) |
|
|
||||
|
No | 162 (58.7) | 50 (18.1) | 64 (23.2) |
|
|
||||
|
5.1 (2) | .08 | ||||||||
|
Graduate and nongraduating | 56 (50) | 28 (25) | 28 (25) |
|
|
||||
|
Undergraduate | 154 (62.1) | 42 (16.9) | 52 (21) |
|
|
||||
|
4.2 (2) | .12 | ||||||||
|
Part-time | 24 (52.2) | 14 (30.4) | 8 (17.4) |
|
|
||||
|
Full time | 186 (59.2) | 56 (17.8) | 72 (22.9) |
|
|
aIncludes 2 participants who identified as nonbinary (queer).
bFisher exact test was used.
Those who used MFDAs at least twice per week reported a 12% lower frequency of consuming 2 servings of vegetables per day compared with those who used MFDAs less than once a week (aPR 0.88, 95% CI 0.63-1.24).
Using “less than once-a-week” as a reference, those who used MFDAs at least twice per week (aPR 0.89, 95% CI 0.57-1.38) and those who used MFDAs once a week (aPR 0.62, 95% CI 0.37-1.38) also reported lower prevalence of not consuming any servings of SSBs per day (11% and 38%, respectively;
Chi-square trend tests revealed no dose-response relationships among frequency of MFDA use, fruit and vegetable intake, SSB intake, and BMI.
Adjusted prevalence ratio (aPR) of frequency of mobile food delivery app (MFDA) use association with diet and BMI (n=360).
Frequency of MFDA use | aPR (95% CI) | ||||
|
.76 | ||||
|
Less than once a week | Reference |
|
||
|
Once a week | 0.95 (0.55-1.65) |
|
||
|
At least 2 times per week | 0.93 (0.55-1.56) |
|
||
|
.24 | ||||
|
Less than once a week | Reference |
|
||
|
Once a week | 0.99 (0.69-1.40) |
|
||
|
At least 2 times per week | 0.88 (0.63-1.24) |
|
||
|
.85 | ||||
|
Less than once a week | Reference |
|
||
|
Once a week | 0.94 (0.67-1.32) |
|
||
|
At least 2 times per week | 1.05 (0.76-1.43) |
|
||
|
.22 | ||||
|
Less than once a week | Reference |
|
||
|
Once a week | 0.62 (0.37-1.04) |
|
||
|
At least 2 times per week | 0.89 (0.57-1.38) |
|
||
|
.09 | ||||
|
Less than once a week | Reference |
|
||
|
Once a week | 1.00 (0.73-1.37) |
|
||
|
At least 2 times per week | 0.91 (0.66-1.25) |
|
aAdjusted for the following variables: age, sex, race, marital status, number of children, current working status, student status, study load, participation in physical activity, BMI, and sugar-sweetened beverage intake.
bAdjusted for the following variables: age, sex, race, marital status, number of children, current working status, student status, study load, participation in physical activity, BMI, and combined fruit and vegetable intake.
cAdjusted for the following variables: age, sex, race, marital status, number of children, current working status, student status, study load, participation in physical activity, sugar-sweetened beverage intake, and combined fruit and vegetable intake.
To complement the survey data, 19 participants were purposefully selected to ask them about their perceptions and experiences of using MFDAs. A total of 84% (16/19) of the interviewees were frequent users.
There were 4 main themes identified from the analyses (see
Conceptual framework that describes the factors that users considered before using mobile food delivery apps (MFDAs).
Before using MFDAs, individuals described considering how they could obtain meals from sources such as inside or outside their homes.
For most interviewees, cooking at home was the most discussed concept. If there were homemade meals, they would not use MFDAs. The action of cooking was mostly carried out by themselves or their parents:
...[I don’t use MFDAs often because my dad]...cook for me...I will just...cook something like vegetables and stuff.
Similarly, if there was no one to cook meals, MFDAs would be used:
...we don’t want to...cook and don’t want to trouble my parents to cook, so we’ll just like order [via MFDAs].
Meals could also be obtained from outside the individuals’ homes. They might dine in, purchase takeout, or even choose pick-up options from MFDAs.
As COVID-19 restrictions regarding attending places of work or study and dining out eased, there were fewer opportunities to use MFDAs:
...because if you’re working outside, you are already outside, so you might as well just go and buy some food from nearby...
MFDA users appreciated the convenience of ordering food. They may need time to finish their work or simply refuse to step out of their homes. Whichever convenience they require justified the use of MFDAs.
Being students, interviewees required time to complete their assignments and revise for exams:
I want to spend more time in study. So I want to save time then I go order.
Similarly, working adults might face time pressure to finish their tasks:
...if it is quite urgent that I...finish the task, I will order online.
Most interviewees were studying and working during the daytime; they would order lunch so that they had more time to complete their tasks of the day. For dinner, they would have more time to prepare their food:
...people are busy during lunch times. They’re either doing work or doing assignments...have some meeting...or rushing from one place to another. At dinner...I have time to...cook...
Staying at home is a convenience. Most individuals did WFH and had web-based lessons for extended periods close to or during the time they were interviewed. These interviewees expressed that they simply did not see the need to step out of their homes for that day as they were already at home:
I’m at home, that’s why I don’t want go out.
As individuals were not stepping out of their homes to procure their meals, there was less probability that users would head out to exercise:
...if...I go and purchase [MFDAs]...I usually don’t do any manual work. Like I don’t do any exercise...So if I just sit here and order delivery...I...feel lazy.
Although young adults understood the importance of the national dietary requirements, they often purchased unhealthy meals from MFDAs:
When they [my friends] order Poke [a salad bowl] [via MFDAs]...I didn’t order Poke with them, and I really appreciated the healthy thing...but for myself, I am highly resistant...I’d like to have something good.
The decision to purchase unhealthy meals related to a sense of self-satisfaction. For many individuals, satisfaction meant relieving hunger:
...[if I am] getting the pizza, I'm getting a garlic bread stick and...a Coke. And if I’m really hungry then I will give myself the concession to go unhealthy and order them...because I know that a salad in the same price [as the pizza combo] will necessitate that I order something else again and again...
For others, satisfaction meant feeling relaxed after stressful events such as work, school assignments, and exams. They would comfort or reward themselves by buying unhealthy options via MFDAs:
I think when I’m stressed...I just want to please myself, I want the instant gratification from the nice food.
Interviewees argued that it was unimportant to adhere to the national dietary recommendation of fruit and vegetable consumption when using MFDAs; rather, it was an individual’s preference to be health conscious:
...I don’t think...when I order, am I really thinking that I need to fulfil the national dietary requirements...I don’t think that’s how I’m thinking while I’m ordering. I’m thinking like...where is good food?
...if you want to eat healthy it’s kind of like an intentional thing. Like you already know that you want to be healthy and you plan for it.
Cost refers to the final cost of the food ordered via MFDAs after the deduction of any prevailing discounts. This theme played the most crucial role in deciding a user’s purchase. If the total cost was low enough, it might override other deliberations described under the preceding themes and have a negative impact on health:
...the delivery fee is much...cheaper...back in China...I used the food delivery every day, maybe for every meal...I was gaining weight...I don’t do the physical exercise...I think I become a couch potato...But in Singapore...the food delivery won’t get so much influence on my health.
This theme was consistent with the observations of infrequent users, who indicated that the main obstacle to MFDA use was the exorbitant price:
I think...the delivery charges and also the cost on the app itself, for certain food items are like higher. So I think that kind of outweighs...the convenience of getting food from the app.
If the cost of ordering from MFDAs was excessively high, frequent users might even decrease their reliance on MFDAs to the point where they became infrequent users:
...back in Indonesia, I rarely cook...I tend to order from delivery app...I think cost...influenced me back then...when the delivery app is...popular in Indonesia, the cost of the food [via MFDAs] is worth a lot cheaper...they [MFDA companies] just want to introduce...so many discounts. But in Singapore, I rarely use the app.
Users would contemplate the total cost of food when choosing food vendors as the food ordered via MFDAs was often expensive to them. Considerations included whether the price of food and the delivery fee were within their budget. If not, they would choose other vendors:
...the first idea is always look at the price of the food. If that’s high...[I’ll]...choose something else. If...the criteria of food price [is fulfilled], the next idea would be...minimum order...let’s say...if I pass,...I see the delivery charge...If it’s reasonably priced...why not?
There were various ways to reduce the costs of food delivery. A commonly used strategy was to search for discounts that offered cheaper food or delivery fees. These discounts may appear in some food vendors, in different MFDA platforms, or in the form of promotional codes. Regardless, these discounts would be attractive enough to determine food purchases:
I look for discounts first before [I] search for the food...see whatever is...cheap delivery.
The framework matrix comparing frequent and infrequent MFDA users revealed that most frequent users could not cook meals for themselves:
I...know...people of our generation don’t really cook.
Instead, frequent users tended to depend on their parents to cook:
...I don’t cook, because...my mother cooks.
In contrast, infrequent users or their parents tended to cook food often. Nonetheless, it is important to recognize that infrequent users might still opt to purchase particular food items from MFDAs if they are incapable of preparing these meals on their own:
...I cannot cook that specific food...for example Mala (麻辣) [commonly used Chinese word to describe for “numbing and spicy” food]...and Bee hoon [rice vermicelli]...and Nasi Lemak [a Malay cuisine] then I will order.
Although not explicitly mentioned, frequent users expressed that they were negatively affected by the COVID-19 restrictions, which resulted in the consumption of extra food or an unhealthy diet:
...[When I was] in quarantine, SHN [Stay-Home-Notice]...[I] start getting hungry at weird times. [I had] some weird cravings...[I] start getting hungry at midnight, even if [I] had lunch and dinner that [the hotel] provided.
...during Circuit Breaker, I’m not as healthy. I ordered bubble tea, I ordered ice cream, I ordered fast-food, I ordered unhealthy food...[because] I feel very sad...[that I] cannot go out...Then [I] eat to be happy lor (a Singaporean colloquial term used to express resignation).
However, infrequent users were not particularly affected by the restrictions.
This study aimed to investigate the use patterns of MFDAs among young adults during the COVID-19 pandemic; examine the association between MFDA use and sociodemographic factors, dietary factors, and BMI; identify the underlying reasons for the observed use patterns of MFDAs among users; and compare the influences of MFDA use between frequent and infrequent users. The quantitative findings showed that 41.7% (150/360) of the participants used MFDAs frequently. Other established use patterns included MFDAs being used more often compared with before the pandemic, lunch being the most commonly bought meal, and Asian-based cuisines and fast food being common food choices. Sex and student status were found to have empirical relevance to the frequency of MFDA use, but this association was found to be not statistically significant. Regression analysis revealed that frequent users had a lower prevalence of consuming 2 servings of vegetables per day and a higher prevalence of drinking SSBs compared with infrequent users, although the differences between frequent and infrequent users were not statistically significant. In total, 4 overarching themes emerged from the qualitative analysis: deliberations about other sources of meals versus ordering in, convenience is vital, preference for unhealthy food ordered from MFDAs most of the time, and cost is king. The framework matrix revealed that the lack of culinary skills and COVID-19 restrictions influenced frequent use.
Taking both the quantitative and qualitative findings into consideration, 4 meta-inferences emerged and will be discussed alongside the existing literature and practical implications. The researchers adopted a narrative-weaving approach that integrates findings from both designs on a concept-by-concept approach [
Meta-inferences derived from the findings.
Quantitative findings | Qualitative findings derived from thematic analysis | Qualitative findings derived from framework matrix | “Fit” of data integration (confirmation, expansion, or discordance) | Meta-inferences |
Most of the participants used MFDAsa more often compared with before the COVID-19 pandemic. | Convenience is vital | COVID-19 restrictions | Expansion | Staying at home is a double-edged sword as it is convenient but can lead to unhealthy behaviors. |
Female participants used MFDAs less frequently than male participants. | Deliberations about other sources of meals versus meals purchased via MFDAs | Lack of culinary skills | Expansion | Cooking is valued by female participants. |
Fast food was one of the most commonly bought meal types. Frequent users were less likely to consume 2 servings of vegetables per day and more likely to drink SSBsb. Graduates and nongraduating students were more likely to use MFDAs frequently compared with undergraduates. | Preference for unhealthy foods ordered from MFDAs most of the time | —c | Confirmation | Young adults’ general indifference toward healthfulness of diet |
41.7% of survey participants used MFDAs frequently. Lunch was the most commonly bought meal. | Cost is king | —c | Confirmation | Cost of food delivery is trivial if time is scarce. |
aMFDA: mobile food delivery app.
bSSB: sugar-sweetened beverage.
cNo suitable findings for data integration.
The main purpose of the COVID-19 restrictions was to prevent the spread of the virus. However, they brought about the secondary benefit of staying at home. It was found that self-interest plays a role in compliance with restrictions [
In the qualitative component, interviewees mentioned that, while doing WFH and learning on the web, they tended to use MFDAs often as they could enjoy the convenience of staying at home, which corroborated the survey findings that individuals used MFDAs more frequently compared with before the pandemic. The qualitative findings also revealed that frequent MFDA users did not feel like engaging in any physical activity outside their homes as COVID-19 restrictions bred sedentary behaviors among young adults [
COVID-19 restrictions had unintended impacts on individuals. Some of the interviewees experienced cabin fever because of staying at home for long durations, which, according to one study, could alter eating patterns [
The qualitative interviews revealed that frequent MFDA users did not cook often. It was suggested in a US study with millennials that they did not have sufficient culinary skills; thus, they preferred to dine in or buy takeout [
Interventions should focus on cultivating cooking habits, especially among young male adults. Before habits are formed, young adults should have confidence in their culinary skills [
Young adults generally have a nonchalant attitude toward healthy eating [
Graduate and nongraduating students were likely to be working adults (
Diet reminders have been proven to be effective in increasing uptake of healthier food choices among people who are actively dieting [
The price of food is a major barrier to using MFDAs according to the interviews. Consistent with the findings of this study, others have found that cost is the single most important determinant of food purchase among young adults [
The interviewees suggested that, although cost is central in decision-making regarding MFDA use, the need for time to finish tasks triumphed over cost. This means that young adults are willing to spend money to buy time to finish their tasks. Young adults had 2 main roles according to the survey: students and employed adults; as such, they may be facing time pressure to finish their assigned tasks, congruent with the findings of a narrative review [
This study has several strengths. First, this mixed methods study used qualitative interviews to explain the quantitative findings in depth and provided clarifications of the phenomenon by triangulating the findings of both designs to explore why young adults reported certain patterns of MFDA use. Second, the use of the framework matrix allowed for the exploration of differences between infrequent and frequent MFDA users, which added further depth to the findings. Third, the interviewees were purposefully sampled with stratification so that the study findings could be more relevant to the demographic profile of Singaporean young adults.
There are several limitations to this study. First, the participants were recruited from only 1 university using convenience sampling, which may not be fully representative of all young adults in Singapore, including those who do not have access to university education. This issue was somewhat alleviated in the qualitative component, in which interviewees were purposefully selected to represent the demographic profiles of young adults more generally and ensure that the perspectives of participants exhibiting high and low MFDA use could be examined. Second, as the response rate was low, there is a possibility of nonresponse bias; nonresponders might provide differing responses from those of responders. Third, BMI reporting in the survey was subjective as survey respondents might have understated their weight to avoid feeling embarrassed, which could have resulted in information bias. Furthermore, the small sample size in the survey may have underpowered the statistical analyses, which may have contributed to the lack of significant statistical associations and the wide CIs found in this study. This limitation was mitigated by the development of the interview guide based on the quantitative results to better explain associations or differences that were not statistically significant. In the qualitative component, XCT was the only coder in this study. As such, there is a possibility that the themes might have been different if there had been a second independent coder. To alleviate this, during the later stages of analysis where latent themes were created, XCT met frequently with NAP to finalize the themes and ensure that they reflected the reality of young adults’ use of MFDAs.
The study’s findings carry considerable implications for the field of public health, particularly in relation to young adults who frequently use MFDAs. These individuals are more likely to have less healthy dietary patterns, with lower vegetable consumption and higher consumption of SSBs. Future interventions could focus on promoting healthy dietary patterns by increasing vegetable consumption and reducing SSB consumption in this population. In addition, interventions could target young adults who are working or learning from home during the pandemic, promoting physical activity and healthy eating habits and reducing sedentary behavior. Moreover, the study highlights the importance of teaching basic cooking skills and promoting cooking as a habit, especially among young male individuals, to encourage healthy eating behaviors. To prevent dependence on MFDAs, it is crucial to educate young adults on effective time management skills. Cost is also a substantial consideration in food purchasing decisions, with this study suggesting that public health policies could explore ways to make healthy food options more affordable and accessible. Furthermore, this study also emphasizes the need for further research to examine the impact of COVID-19 restrictions on young adults’ dietary patterns and physical activity levels as well as evaluate the effectiveness of interventions aimed at promoting healthy behaviors during periods of social distancing and restrictions and during the
Quota sampling frame.
Questionnaire.
In-depth interview guide.
Univariate analysis of the associations between frequency of mobile food delivery app use and fruit and vegetable consumption, sugar-sweetened beverage consumption, and BMI.
Summary of themes, subthemes, and other examples of evidence.
adjusted prevalence ratio
Checklist for Reporting Results of Internet E-Surveys
mobile food delivery app
noncommunicable disease
National University of Singapore
Research Electronic Data Capture
sugar-sweetened beverage
Saw Swee Hock School of Public Health
work from home
The authors would like to express their gratitude to the members of National University of Singapore Saw Swee Hock School of Public Health Physical Activity and Nutrition Determinants in Asia Team for their valuable contributions to this study, which included providing helpful suggestions and reviewing both the questionnaire and interview guide. This study, conducted as part of the Master of Public Health graduation requirement, did not receive any external funding.
The data sets generated or analyzed during this study are available from the corresponding author upon reasonable request.
XCT, FM-R, and NAP contributed to conception and design of the study. CW provided her subject matter expertise in this study. XCT contributed to data collection for both the quantitative and qualitative designs. XCT organized the database and performed the statistical analysis with critical review from NAP and FM-R. XCT transcribed the audio recordings and performed thematic analysis with critical review from NAP. XCT wrote the first draft of the manuscript with critical revision from NAP. All authors contributed to manuscript revision and read and approved the submitted version.
None declared.