Effectiveness of Integrated Technology Apps for Supporting Healthy Food Purchasing and Consumption: A Systematic Review

A healthy diet is essential for good health and nutrition, though literature showed that there are various factors affecting the intention to purchase and consume healthy food. Technology integration is known to be useful in various aspects, but findings from studies on the efficacy of technology integration to improve healthy food consumption and purchase have largely been inconsistent. Therefore, we aimed to examine the efficacy of interventions that use technology apps to improve healthy food purchasing and consumption in adults. Relevant studies were identified through PubMed, Scopus, CINAHL, SportDiscuss and ACM Digital Library. Twenty studies were included in the systematic review. The majority of studies (n = 18) used a smartphone in the intervention, and only two studies used a personal digital assistant. The results showed that technology integration-based intervention favoured healthy changes in household food purchases, and increased consumption of healthy food and healthy eating outcomes – albeit to different extents. Overall, technology apps are convenient and user-friendly tools to encourage a change in healthy food purchase and consumption among people.


Introduction
Healthy food purchases and consumption are the trending topics among consumers nowadays. An increase in healthy food purchasing patterns may motivate consumers to practice healthier food consumption habits [1]. While the definitions of "healthy" foods differ, generally, nutrition standards emphasising healthier foods would decrease the levels of sodium, sugar and unhealthy fats, while increasing the intake of fruits and vegetables, whole grains and lean protein [2]. On the other hand, unhealthy food consumption, such as consuming ultra-processed food, high calories and high-fat food, is a proven risk factor for a variety of chronic diseases that cause millions of deaths and disabilities each year globally [3][4][5][6][7][8]. Previous studies have showed that the prevalence of overweight and obesity among type-2 diabetes patients was 36.2% in Southwest Ethiopia; while 28.1% and 5.2% of overweight and obesity adults respectively in China had associations with factors such as food consumption, gender, income level, family history, physical inactivity and marital status [9][10][11]. Consequently, encouraging healthier food purchasing and consumption is critical for population health.
In recent years, around 75 percent of individuals have used various types of mobile interventions [12], and the mobile phone has become an essential medium of communication throughout the world [13]. Apps on mobile devices, such as applied on sensors and mobile apps, are becoming the centre of customers' digital lives, and they are widely used. Many interventions are linked to the widespread usage of smartphones and apps nowadays. Apple's iPhone, the first modern smartphone with the ability to add apps, was released just over ten years ago, in 2007 [14]. The smartphone, which is essentially a computer in your pocket, has become both personalised and easily accessible in recent years. As eating is an everyday activity, the prospect of altering behaviour using digital technologies in general is very enticing. Therefore, the use of technology apps intervention could be a promising approach towards altering people's behaviour towards becoming a healthier consumer.
There has been an increase in the number of studies that have used technology interventions to alter people's behaviour toward becoming healthier [15][16][17][18][19]. Technology-based behavioural health interventions involve the delivery of evidence-based practises via text messaging, apps, social media and multiple health components [20][21][22], through platforms such as computers, mobile phones or wearable sensors [23]. Due to rapidly growing evidence of their effectiveness and efficiency, technology-based behavioural health interventions are gaining traction as therapeutic resources, both as stand-alone technology or as multi component interventions [24][25][26][27][28][29][30][31][32][33]. Reflecting on this trend, there are reviews on the efficacy of stand-alone technology apps intervention or multicomponent intervention on healthy lifestyles [34][35][36][37][38][39][40][41], but still, the developing literature base on technology-based intervention has primarily focused on disease prevention [29,38,42,43], weight management [15,16,18,44,45] or lifestyle improvement outcomes [17,19,44,46], and in many ways has not yet explicated the important implications for a stand-alone technology intervention on healthy food purchases and healthy food consumption outcomes. Along with the relative lack of attention, specifically on healthy food purchasing and consumption, this review paper poses the following important research questions: Are technology app-based interventions effective enough to support healthy food purchase and consumption when compared with the traditional interventions such as face-to-face counselling, literature and group-based education? In which form are technology apps more effective and commonly used by adults? Is the effectiveness of stand-alone technology apps intervention comparative with the multi-component intervention? Which group of people have more access to technology apps? In order to fill in the knowledge gap, the general purpose of this review is to examine the efficacy of interventions that use technology-based apps to improve both healthy food purchasing and consumption in adults. Specific objectives were: (1) to determine the effectiveness of technology apps-based intervention to support healthy food purchase and consumption; (2) to identify the types of technology apps that are more effective and commonly used by adults; (3) to determine the potential of efficacy of stand-alone technology apps intervention on healthy food purchasing and consumption; and (4) to identify the target group of people who have more access to technology apps.

Materials and Methods
The study methods were registered in the international prospective register of systematic reviews (PROSPERO) protocol, with the registration number: CRD42020218742 [47]. We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines when conducting this systematic review and reporting its results [48].

Searching Strategy
The electronic databases searched were PubMed, Scopus, CINAHL, SPORTDiscuss and ACM Digital Library. These databases were searched for studies published in journals between 1 January 2006 and 31 December 2020, as it was considered unlikely that apps interventions were developed before 2006/2007, when smartphones were first introduced. The search was limited to English-language publications, and restricted to studies focusing on apps on food purchasing and food consumption in adults. Table 1 presents the search terms used in the present review.

Search Category Search Terms
Apps Application* OR app* OR smartphone* OR "smart phone*" OR "cellular phone*" OR "mobile phone*" OR "mobile telephone*" OR tablet* OR "e-learning" OR "e-health" OR "iPad*" OR "mobile health" OR "social media" Food purchasing and Food consumption "food purchasing" OR "food purchase*" OR "food choice*" OR "food planning" OR "food shopping" OR "food consumption*" OR "food intake*" OR "dietary intake*" OR "healthy eating" OR "eating behaviour*" OR "healthy food" OR "food behaviour*" OR "food value*" OR "food diary" OR "food diaries" OR "nutrition assessment*" OR "diet record*" OR "diet survey*" OR "energy intake*" OR "nutrition survey*" OR "dietary assessment*" Intervention Intervention* OR program* OR programme* OR "health promotion*" OR trial* OR effectiveness Table 2 shows the domains, inclusion and exclusion criteria. The inclusion criteria were: (1) Publication years between 1 January 2006 and 31 December 2020; (2) original articles published in peer-reviewed journals; (3) studies in the English language; (4) studies from any country and any population; (5) targeted adults who are aged 18 and above; (6) quantitative studies with experimental or cross sectional study design that reported technology apps as an intervention group; and (7) studies must contain the food purchasing or food consumption as the outcome.

Study Selection
First, two investigators imported the potential articles from the databases into the Endnote Program X5 Version separately. Any duplicate publications were removed by Endnote Program X5. A manual check of the redundant publications was performed by the two investigators. Second, two investigators independently screened the titles and abstracts of those articles for suitability based on the search strategies mentioned above. Third, full-text articles were independently assessed based on the inclusion and exclusion criteria. Lastly, all articles that were agreed upon after discussing them with the third investigator were included. Studies published before or after the inclusion dates.

Publication type
Original studies published in peer-reviewed journals only.
Letters, commentaries, conference proceedings, reviews, narrative articles or other materials that was not a peer-reviewed primarily study.

Language
Studies limited to English-language publications. Studies were not published in English.

Targeted population Any population.
No restriction on population.

Targeted group
Adults who are aged 18 or above. Children or adolescents below 18 years old. Qualitative study or any non-experimental study designs were utilised (e.g., protocol, studies reporting prevalence or trend data, feasibility studies, measurement studies or theoretical papers).

Study scope/variables
(1) Technology apps [e.g., smartphone or personal digital assistant (PDA)] in an intervention to improve healthy food purchasing and consumption for prevention.
(2) Interventions could be stand-alone interventions using an app only, or multi-component interventions including an app as one of several intervention components (e.g., physical education, face-to-face counselling) with the condition that individual apps record was provided.
(3) All types and units of measurements for the healthy food purchasing and consumption outcomes were acceptable (e.g., food group, self-report, servings, calories, kilograms, nutrition assessment).
(4) reported data regarding efficacy for behaviour change (e.g., change in healthy food groups). (5) Interventions covering aspects of large-scale management of food (retail, restaurants, public preparation and consumption of food in school kitchens, hospitals, etc.).
(1) tracking app only but not focused on apps intervention on food purchasing or food consumption.
(2) Other technology intervention than apps intervention (e.g., text messaging, apps, social media, telephone counselling or online coaching).
(3) No measure of healthy food purchasing or food consumption outcomes.

Data Extraction
The following data were extracted by two investigators independently using a standardised data extraction template with the information such as the first author, year of publication, country, study name or app name, device, study design, sample characteristics, number of participants, mean age, percentage of females, grouping, intervention time frame, outcomes and measures and the finding of the studies. The results of data extraction were compared between the two investigators to ensure no errors.

Quality Assessment
We used Transparent Reporting of Evaluations with Nonrandomised Designs (TREND) to assess the quality of included studies [49]. This tool consists of 46 items that assessed components in non-randomised trials. Whenever the information provided was not enough to assist in making judgement for a certain item, we graded the item with a "0" meaning absence of component, and "1" was awarded to a particular item if information on a particular component was sufficiently described. Each article's quality was graded as "Good" if the TREND score ≥37 or equivalence to 80%; or "Poor" if TREND score was <37. In this review, we would not exclude any studies based on their TREND score. The detail of scoring results is shown in Supplementary Table S1.

Data Analysis
All authors read, understood and synthesised the findings. Several discussions were conducted among the team to categorise findings into the following themes: "efficacy"; "technology apps"; "unhealthy foods"; and "healthy foods". We follow the definition as below: • Efficacy was defined as the ability to produce a desired or intended result [50]. • Technology apps were defined as applications for novel mobile consumer devices with a touchscreen, especially smartphones and tablet PCs [51]. In this review, the terms "app" is sued in the stricter sense with more on self-monitoring apps. Other technology interventions such as text messaging, social media (e.g., facebook), online coaching or telephone counseling were excluded in this review. • Unhealthy foods were defined as those that were high in salt, sugar and saturated fats [2]. • Healthy foods were generally defined as the food meeting the nutrition standards promoting healthier foods that were low in sugar, salt and saturated fats, while promoting fruits and vegetables, whole grains and lean protein [2].

Description of Included Studies
A total of 4175 articles were identified in the initial screening, as shown in Figure 1. After the removal of duplicate articles (n = 824), a total of 3351 articles were retrieved for the review of their title and abstract to determine if inclusion and exclusion criteria were met. Of the 3351 abstracts, 147 studies were identified that needed a further full-text review. After the careful evaluation of the 147 articles, 20 articles were included in this systematic review.

Characteristics of Included Studies
The characteristics of the included studies are summarised in Table 3. In general, most of the studies were conducted in the United States (n = 9, 45%) [24-26,30,32,52-55],

Quality Assessment
The summary of the quality assessment is presented in Table 3, and the detail of scoring result was shown in Supplementary Table S1. The TREND score ranged from 28 to 39 (total mark is 46), or equivalent to 60.9% to 84.8%. Out of 20 included studies, four studies were marked as "good" quality and 16 studies remarked as "poor" quality. The most common absent component (none of studies reported these components) seen across included studies were "8c: Inclusion of aspects employed to help minimise potential bias induced due to non-randomisation (e.g., matching), 12b: Description of protocol deviations from study as planned, along with reasons", 14c: Baseline comparisons of those lost to follow-up and those retained, overall and by study condition" and "17c: Inclusion of results from testing pre-specified causal pathways through which the intervention was intended to operate, if any".

Healthy Food Purchasing and Its Measurement
There are two studies that specifically showed the household food purchases [29,54]. Eyles et al. [29] aimed to determine the effectiveness of a technology app to support people with cardiovascular disease to make low-salt food choices [29]. The primary outcome of the study was the salt content of household packaged food purchases during the four-week intervention (g/MJ). Secondary outcomes were the saturated fat content (g/MJ), energy content (kJ/kg) and expenditure (NZ$) of household food purchases. Results showed a significant reduction in mean household purchases of salt, but were not significant on the saturated fat content (g/MJ), energy content (kJ/kg) and expenditure (NZ$) of household food purchases, in the intervention app group as compared with the control group.
On the other hand, Palacios et al. [54] aimed to pilot test the effectiveness of a smartphone app that generates healthy grocery lists, on the indices of diet and weight. Results showed a significant improvement in the app group compared to the baseline, but no significant improvement compared to the traditional group (face-to-face counselling sessions) [54]. Although other included studies did not specifically measure the healthy food purchase, the measurement of healthy food choices may correlate with healthy food purchases. The summary of the healthy food purchase results is showed in Table 3.

Healthy Food Consumption and Its Measurement
The measurement of healthy food consumption varied across all studies. The measurement used included food choice (varies from type of food: vegetables, fruits, whole grains, sugar-sweetened beverage, unhealthy snacks, fatty food and alcohol consumption), dietary intakes (measured by calories, energy, total fat, carbohydrates, protein and fibre intake), nutrients intake (macronutrient: carbohydrate, fat and protein, or micronutrients: vitamins and minerals intake or fibre consumption), healthy eating index (HEI), healthy eating behaviour and healthy dietary behaviour. The instrument to measure food intake included self-reported or face-to-face interview using two-day food records, food frequency questionnaire (FFQ), 24-h dietary recalls or self-administrated questionnaire.
Some researchers collected food consumption data for other purposes of analysis. For example, some researchers used apps to monitor the dietary intake and assist in weight management [24,26,59]. Dodd et al. (2017) used apps to provide lifestyle advice to pregnant women [56]. Gill et al. (2019) aimed to determine the influence of an app to improve healthy eating to reduce modifiable risk factors for chronic disease [58]. Huberty et al. (2019) aimed to test the efficacy of a mobile app on stress, mindfulness and self-compassion in college students [55]. Other published papers aimed to test the efficacy of apps in changing healthy lifestyle among patients [28,33,53]. The summary of the healthy food consumption results is showed in Table 3.

Intervention Efficacy
A summary of intervention effects for the included healthy food purchase and healthy food consumption outcomes (macro and micronutrients, healthy food types intake, healthy eating index and healthy eating behaviour) are presented in Table 4. Overall, only two studies discussed the healthy food purchasing as a outcome [29,54]. Both studies showed no significant difference between apps and the control group for healthy food purchasing, except Eyles et al. (2017), which demonstrated the finding that intervention group chose lower salt content when making food purchases.
A slightly larger proportion of studies reported vegetables (15 out of 21; 71%), fruits (14 out of 21; 67%) and grains intake (6 out of 21; 29%) as healthy food consumption. Only two studies showed significant improvements in vegetables [27,52] and fruits [27] intake in favour of the app intervention group. However, Rodgers et al. (2015) also showed that participants with a BMI < 21 kg/m 2 have lower consumption of fruit intake in the app intervention group. Two out of 12 studies showed improvement in lower fatty food consumption in the intervention group [19,25]. One out of six studies had findings of more grains intake in the intervention group [52]. For the unhealthy food choices such as added sugar, salt and unhealthy snacks, there is no significant difference between the intervention group, nor within intervention and control group, except one study that showed that the intervention app group has lower unhealthy snack consumption than that of the traditional group (face-to-face counselling session) [54]. Furthermore, 50% of the studies demonstrated a positive result i.e., app intervention improves healthy eating behaviour [31,32,58].

Discussion
This systematic review found modest evidence for the efficacy of app interventions to improve healthy food purchasing and food consumption. The majority of these studies found that significant between-group improvements in the apps intervention group versus the comparison group, which is the gold standard for evaluating the efficacy of health interventions [61]. Despite the limitations of the studies, the results of this review suggest that apps can be an important tool for improving healthy food purchasing and healthy food intake. The effectiveness of apps interventions can be explained in part by the advantages of mobile applications over other intervention delivery modes such as text messaging, websites, paper journals, email, face-to-face counselling and group sessions. Since many people lead busy lives, they may prefer applying easier access to health programs using smartphone or PDA apps anytime and anywhere, rather than spending time on looking for the paper journals, or making appointment for the face-to-face counselling or group sessions at particular times. Besides, the well-developed apps may provide the sufficient information and guidance, real-time self-monitoring, feedback, motivation, social support and incentives when they achieve the goals, in assisting them with health behavioural changes [62]. In this sense, technology apps are effective tools in this era where most people are looking for convenient, cost and resource savings interventions.
Among the included studies in this review, since 2006, when the integrated applications were launched, hand-held computers were utilised to improve and monitor the dietary intakes of the consumers [25,52]. Only until recently, smartphone applications were utilised in obesity treatment to achieve weight loss among smartphone users [24,26,57,60]; as well as in salt reduction movements, through buying low salt content food purchase behaviours [29]. Western countries were still the leading countries for the technology app penetration when compared to Asian countries. During the pre-and post-COVID-19 pandemic, smartphone applications were largely adopted among consumers due to their accessibility, trust, user friendliness, satisfaction and great performance, with almost half of consumers (45.68%) using the apps at least once every three days [63]. During the COVID-19 pandemic era, many countries were locked down, and the mandatory closure of gyms and outdoor recreation venues (such as parks, gardens, swimming pools, basketball courts and stadiums) may limit one's movement in their own house, and consequently cause increasing levels of sedentary lifestyle among the population. Health-related issues have been in rather alarming increments, and technology apps may be an effective tool to engage people in healthy behaviours. Hence, the integrated technology applications in embracing both healthy food purchase and consumption were considered practical solutions to prevent health problems such as obesity and other non-communicable diseases, while promoting healthy food purchase and consumption at the same time, akin to fixing two major health concerns simultaneously. Technology-based apps are expected to be used by more people globally regardless of geographical locations in the very near future. In recognising this need, many governments around the globe, including in Malaysia, have subsided the provision of cheaper internet, or even smartphones, to enable people to remain connected, especially with regards to health-related issues.
Notwithstanding the potential of the apps, most of the intervention studies included in this review showed the significant findings only in selected food intakes, but not all, which means that the comparison group is still playing its role in health intervention. Technology apps are a motivational and self-monitoring tool, but it is highly dependent on the participant's motivation and intention to use the app. Studies suggested that more intensive follow-ups with apps notifications may enhance the efficacy of the apps intervention [54,59]. To counter this, technology apps must provide the user-desired needs of the combination of credentialed knowledge, interactivity, personalisation, and individual feedback, in addition to providing information about positive food-related behaviour changes [64,65]. On the other hand, the insignificant findings of the between-group improvements of some of the studies also raises the question whether multi-component interventions have a greater impact than single-app interventions. To achieve long-term health behaviour improve-ment, several reviews of health behaviour change programmes suggest using multiple intervention techniques to achieve better outcomes [66][67][68][69]. As a result, it is possible that integrating apps into multi-component interventions yields better health outcomes than stand-alone app interventions; however, future studies will need to confirm this.
All of the included studies showed that younger aged persons formed the majority of participants in the intervention apps group, and the majority of participants were females. It demonstrated that age and gender affected smartphone usage, and also contributed to the significance in results. It is supported by research which reported that younger adults use their phones for longer periods of time, and their usage is primarily directed toward entertainment and social interactions via specialised apps [70]. Older participants use them less frequently, and primarily for information or as a traditional phone [70]. Furthermore, females use smartphones for longer periods than males, with an average of 166.78 min vs. 154.26 min per day, respectively [70]. In terms of purchasing behaviours, female purchasers opted to choose healthier food options such as vegetables and fruits, when compared to male purchasers [71]. Women from lower socioeconomic income groups, on the other hand, were considered as purchasing less healthy food choices that did not manage to meet the recommended dietary intakes for vegetables, milk and oils [72]. The individuals from the lower socioeconomic income group also possessed higher intention to consume oversized portions of unhealthy food, when compared to those from the higher socioeconomic group [73]. With this finding, technology app interventions should be more effective for younger adults and female users. Future designs for app development should be based on differences between the different groups of individuals.
Only two studies evaluate the effectiveness of apps on healthy food purchasing, showing improvements within the apps intervention group, in selected food categories [29,54]. However, two particular studies were directed at different health behaviour outcomes such as weight management [54] and lower salt intake for cardiovascular disease patients [29]; this is evidence that the existing technology apps clearly lack a focus on healthy food purchasing in improving general health. Given the importance of food purchasing behaviour as an important step in the food consumption process [74], as well as the potential effectiveness of mobile apps to support behaviour change, there is a need to further investigate the potential role of existing mobile apps.

Strengths and Limitations
This review synthesises the existing literature and identifies any remaining gaps. It generates a comprehensive summary and discussion on the topic about the technology apps intervention on healthy food purchase and healthy food intake. The research characteristics, study designs, intervention durations, outcomes and findings are all covered in this review. Based on our analysis, a number of gaps in this field were identified, as highlighted in the discussion section above.
Another strength of this review is that it only identifies the efficacy of technology apps on food purchasing and food intake. It extracts the technology app intervention findings from multi-component interventions. Moreover, it draws on the food purchasing and food intake data from the studies which are not focused on food purchases or food consumption alone. For instance, studies focusing on weight loss, mindfulness meditation and lifestyle modification for a targeted group of patients were included in this review.
Potential biases may have affected the findings of this review, which is one of its limitations. First, limiting this review only to English language-based studies as one of the selection criteria may cause a selection bias. It introduces the risk of ignoring data that is not published in the English language, and may cause missing out on important cultural contexts which are tied to geographical aspects; this may therefore limit the review's findings. However, a systematic review has demonstrated no evidence of systematic bias from the application of language restrictions on choosing the articles [75]. To assess the effect of language restrictions on systematic reviews in health sciences, further research is required. Secondly, publication bias could be present, which could be due to the lack of negative effects of the reported interventions in this review. Lastly, we found that the interventions mentioned in older studies from 2013 vary from those described in newer studies. Technology apps with older interventions may not be as effective as newer apps with advanced technology such as food photo recognition and personalised realtime feedback.

Conclusions
In conclusion, technology apps are an effective media, and are comparative with the traditional interventions for improving healthy food purchasing and nutrition-related outcomes among many diverse communities, especially among younger adults and female users. Smartphone apps are more effective, accessible and commonly used by adults, and can be used for stand-alone technology app intervention, which has the comparable positive effect with the multicomponent intervention. The strengths of technology apps are attributed to their nature of low cost, wide reachability and people having more time available to browse for online purchasing platforms, especially during the COVID-19 pandemic. However, the need for continued engagement and certain complicated app features are the key factors that can limit its effectiveness. Therefore, app development should be tailored with a different degree of healthy food choices. There is definitely more room for development and testing, particularly in this ever-vibrant app market, to establish its long-term efficacy.  Data Availability Statement: Data from this study is contained within this article and supplementary file.

Conflicts of Interest:
The authors declare no conflict of interest.