Correlations between Convenience Cooking Product Use and Vegetable Intake

Australians’ vegetable intakes are low, and strategies are needed for improvement. Popular convenience cooking products (meal bases and recipe bases, ready-made marinades, and convenience cooking sauces) address common cooking and vegetable consumption barriers (cost, time, and cooking skills). However, relationships between their usage and vegetable intakes have not been established. Therefore, Australian adults were surveyed on convenience cooking product use, vegetable intake and variety, behaviours when barriers to vegetable inclusion arise, and vegetable choice factors. Of 842 participants, 36.7% used meal and recipe bases, 28.1% marinades, and 47.2% cooking sauces, with most following the back-of-pack recipes at least sometimes. A total of 12.5% of participants used products from all three categories. Factors associated with lower vegetable intakes were meal and recipe base and cooking sauce use, using a higher number of product categories, and always following back-of-pack recipes. Factors associated with lower vegetable variety were the use of meal and recipe bases and cooking sauces. Factors in vegetable choice, and behaviours when not including a listed vegetable (due to not having or liking the vegetable, or an inability to eat it) did not vary by usage habits. These results provide insights into current vegetable intakes of those using convenience products, providing a baseline for future changes in the product design and recommendations.


Introduction
It is well established that diets regularly containing a high level and variety of vegetables are beneficial to overall health, reducing risks for diet-related chronic diseases [1][2][3]. As such, the World Health Organization recommends consuming 400 g of vegetables per day [4]; similarly, the Australian Guide to Healthy Eating suggests that most adults eat a minimum of 375 g of vegetables a day (at least five serves of vegetables at~75 g per serve) [5]. Recommendations also specify that intake should include a variety of vegetables, as each vegetable has a unique profile of nutrients and other healthful compounds [6]. Nevertheless, the results of the 2018 Australian National Health Survey demonstrated that~95% of participants did not meet these recommendations [7]. Importantly, despite multiple public education campaigns, this figure had not changed over the ten preceding years [8].
There is an array of barriers that can contribute to low and limited vegetable intake [9]. These include a lack of availability, time, and access, as well as burdens of cost and taste preference [10][11][12][13]. It has been noted that, in the broader population, cooking skills have been at a decline, and devalued, which is a concern for individual food choices and health outcomes [14]. Importantly, vegetables often require preparation or cooking prior to

Study Desgin and Recruitment
This cross-sectional survey (via Qualtrics, SAP, Provo, UT, USA) was conducted using snowball recruitment, as described in Brasington et al., for~7 weeks in 2020 [22]. Participants were living in Australia, over 18 years of age, and proficient in English comprehension skills, as the survey was administered in English. Ethics approval was granted by the Human Research Ethics Committee of the University of Newcastle (Reference No. H-2020-0119). The online questionnaire used both qualitative and quantitative questions developed by the authors, and was piloted internally. Questions were arranged in thematic blocks based on consumption of vegetables, factors determining vegetable choices, convenience cooking products usage habits, and demographics. The participants were asked to self-identify if they used convenience cooking products ("Do you use any of the following products?") with several product types listed (to ensure recognition across different brands) and collapsed into the following categories for analysis: meal and recipe bases (including meal bases, recipe bases, and recipe concentrates), readymade marinades (marinades only), and convenience cooking sauces (including simmer sauces, pasta sauces, and other sauces), or none of the above. The number of product categories used was also summed. Participants were then categorised as "users" or "nonusers" of these products. Regarding back-of-pack recipes, participants were asked if they always, sometimes, or never follow the recipes provided with products [22]. In a matrix style question, users were also asked to select from a list of actions what they were most likely to do if they were following a back-of-pack recipe on the product classes they reported using and either did not have, did not like, or could not eat a vegetable listed (not including  that vegetable, replacing it with a different vegetable, replacing it with a similar vegetable,  not making the meal at all, not sure, other).
Participants were asked to report their typical daily vegetable consumption in serves per day (with a serves guide provided) using the question "How many serves of vegetables do you typically eat per day (1 serve of vegetables is half a cup of cooked vegetables or 1 cup of salad)?", and a numerical dropdown provided. Responses were accordingly categorised into either meeting or not meeting the daily recommendation for vegetables as per the Australian Guide to Healthy Eating. As an indicator of vegetable variety and consumption, eating frequency of fourteen of the most commonly consumed vegetables in Australia was assessed, based on the Australian Healthy Eating Quiz [27]. Participants were then given a vegetable variety score (score 1 for each vegetable they consumed at least one serve of regularly, i.e., at least weekly), with a possible range of scores from 0-14.
Factors involved in determining vegetable selection were rated on a 5-point Likert scale (ranging from 1 = not at all important to 5 = extremely important). There were 12 factors rated (taste, flavour, costs, sustainability, availability, texture, colour, nutritional content, easy to cook, shelf life, quality [28], and value for money).

Statistical Analysis
Data were analysed using the statistical analysis software package JMP (Pro 14; SAS Institute Inc., Cary, NC, USA). The threshold for statistical significance was set at a p-value of <0.05, and p-values were reported to one significant figure. Contingency tables (Pearson χ 2 ) and nominal logistical regression were used to assess the differences in distributions between categories. T-tests and ANOVA with Tukey's post-hoc tests were used to compare the differences in continuous variables between groups. It was hypothesised that those who used convenience cooking products would have lower vegetable intakes and variety scores than those who never used them, but that the following of the recipes would be linked to higher vegetable intakes and variety scores.

Results
A total of 964 people gave informed consent, and participated in the survey. One hundred and twenty-two of those participants were excluded due to incomplete responses, or for having completed the survey in less than half the median completion time (244 s), or for failing the attention check (similar questions with reversed scales). Overall, there was a total of 842 responses included in the final sample.
The age of respondents ranged from 18-80 years (median 41 years, standard deviation 12.1 years). The sample was mostly female (77.7%), and the majority had a university level of education ( Table 1). The majority had household incomes above AU$75,000 per year (equivalent to approximately €47000), and reported working full-time hours (30 h per week or more; Table 1). 36.7% of participants reported using meal and recipe bases, 28.1% used marinades, and 47.2% used cooking sauces (including pasta sauces and simmer sauces). There were no differences in incomes, education, work hours, or sex distributions between the user and non-user groups for any of the convenience cooking products. Those who used convenience cooking sauces cooked at home less frequently than non-users (p = 0.003, Table 1). Regarding the number of individual convenience cooking product classes used, 41% of participants used none, 23.5% used only one product type, 22.5% used two product types, and 12.6% used all three.

Following of Back-of-Pack Recipes
The majority of participants who used meal and recipe bases, marinades, and cooking sauces followed the recipes provided on the back of the pack at least sometimes (sometimes (41.9-44.1%) or always (21.9-24.9%)); however, 32.7-34.2% of participants never followed the recipes provided (Table 2).

Reported Vegetable Intake and Variety Scores
Mean vegetable intake in the sample was 3.1 serves per day (standard deviation = 1.2 serves per day). Users of meal and recipe bases were less likely to meet the daily recommended vegetable intakes than non-users (10.4% vs. 17.3%, respectively: χ 2 = 7.4, p = 0.006). Similar results were found for users of convenience cooking sauces (10.8% vs. 18.2%; χ 2 = 9.0, p = 0.003). However, there were no differences in the proportions meeting the recommended thresholds in ready-made marinade users compared to nonusers. When reported daily vegetable serves were considered as a continuous variable, those who reported using any of the convenience cooking products (meal and recipe bases, marinades, or cooking sauces) had lower reported mean daily vegetable intakes compared to non-users (p ≤ 0.005, Table 3). However, there was no significant difference between vegetable variety scores between users and non-users of any product class (Table 3). Among meal and recipe base users, those who reported never following the recipes were the most likely to meet the daily recommended intakes (16.0%) compared to those who reported always (5.9%) and sometimes (6.9%; χ 2 = 8.0, p = 0.002) following the recipes. Similar results were found for users of convenience cooking sauces (never = 15.4%, always = 4.35%, and sometimes = 9.2% meeting recommended intakes; χ 2 = 8.8, p = 0.01). However, there were no differences in the proportions meeting the recommended thresholds by recipe-following for marinade users (p = 0.1). When reported daily vegetable serves were considered as a continuous variable, meal and recipe base users who reported always following the recipes had lower reported mean daily vegetable intakes compared to sometimes and always followers (p = 0.0001, Table 4). Users of meal and recipe bases who reported always following the recipes also had lower vegetable variety scores than those who never or sometimes followed (p = 0.0001; Table 4). For convenience cooking sauces, those who always followed the recipes had lower vegetable intakes and variety scores than those who sometimes followed (p = 0.0008 and p = 0.009, respectively; Table 4). However, intakes and variety scores did not differ by recipe-following for marinade users (p = 0.7; Table 4).
Those who did not report using any of the convenience cooking products had higher vegetable intakes than those who used two or three product classes (Table 5). However, vegetable variety score did not vary by number of product categories used (Table 5).  Those who did not report using any of the convenience cooking products were the most likely to meet the daily recommended intakes of vegetables (19.5%) compared to those using one product class (11.1%), two product classes (13.8%), or three product classes (7.6%; χ 2 = 12.8, p = 0.005).

Decisions When Barriers Arise to Including the Vegetables Specified on Back-of-Pack Recipes
Users of convenience cooking products were asked what they would do if they did not have, could not eat, or did not like vegetables listed in the recipes provided with meal and recipe bases to assess their behaviours when faced with these barriers. The vast majority of participants reported that they would replace that vegetable with another vegetable or a similar one if they did not have (87.4-88.7%), did not like (85.2-89.0%), or could not consume (81.9-87.7%) the vegetable listed (Table 6). Not including the vegetable and not making the meal at all were the least common courses of action. These results were similar when analysed by number of convenience cooking product categories used (Table 7).

Factors Determining Vegetable Choices
The top three determinants of choice regardless of convenience cooking product use or number of categories of product used were taste, flavour, and quality (Tables 8 and 9). The three least important determinants were shelf-life, sustainability, and colour (Tables 8 and 9). There were no significant differences in the mean levels of importance given to each factor in determining vegetable choices between those who reported using each convenience cooking product and those who did not, other than a small difference in the importance score for shelf-life in convenience cooking sauce users compared to non-users (Table 8), and sustainability between the number of convenience cooking products used (Table 9).   Table 7. Actions taken when barriers to including a vegetable listed on a convenience cooking product back-of-pack recipe arises, by number of categories of convenience cooking product used.

Discussion
This study is the first to investigate the relationships between usage habit key examples of convenience cooking products, and vegetable consumption habits. It appears that the use of these products, both in binary terms (user relative to non-users) and in terms of using a higher number of convenience cooking product categories, was associated with lower vegetable intakes. This may reflect the fact that those who choose convenience cooking products face more barriers to obtaining a healthy diet and accessing vegetables. We have previously demonstrated that those who use convenience cooking products (including meal and recipe bases, marinades, and other sauces) have lower cooking confidence and creativity [22], and it is established that those with lower cooking skills are less likely to eat a healthy balanced diet [29][30][31]. Additional research is needed to determine if convenience cooking products are acting as a tool to replace other less healthful convenience foods, such as takeaway foods and ready-made meals accessed outside of the home, or if they are displacing more healthful home-cooked meals, in order to assess the utility of these products in encouraging a healthy diet.
Furthermore, those who used meal and recipe bases and other cooking sauces, and always followed the recipes provided on the back of the pack had lower vegetable intakes and variety scores than those who never followed the recipes. Again, this may be linked to cooking confidence and creativity, with those with low creativity and confidence more likely to follow recipes. However, we have also previously conducted an audit of the back-of-pack recipes provided with meal and recipe bases available at major Australian supermarket chains, and demonstrated that these products are typically low in total serves of vegetables, and are low in vegetable variety, featuring mostly starchy and orange vegetables, and low levels of green vegetables [26]. Therefore, reformulation of these back-of-pack recipes to include more vegetables, or to recommend side salads, could help to improve vegetable intake in those who use these products and follow the recipes. These associations were not found for users of ready-made marinades, and as such, more investigation may be needed to distinguish between market segments of different convenience cooking products.
Importantly, though a total of 76.2-78.1% of participants reported only following the recipes sometimes or never, the vast majority of participants reported that when they did not have, could not eat, or did not like a vegetable ingredient listed, they would replace that ingredient with another vegetable (either a different vegetable or similar). This means that manufacturers can be confident that consumers are obtaining the serves of vegetables listed in the recipes when designing these recipes to encourage vegetable intakes. The deviation to different vegetables may also explain why those who do not always follow the recipes have higher vegetable variety scores.
Though time, knowledge, and cost are commonly cited barriers to vegetable consumption [9][10][11][12][13][14][15][16], income, education, work hours, and selection factor variables did not vary by usage or recipe-following habits. However, it is important to note that this sample was collected as a snowball sample and convenience cohort, resulting in a selection bias toward women with high incomes and education levels, as is typical for this method [18]. As such, these findings require validation in a more balanced sample. However, women remain disproportionately responsible for home cooking and grocery shopping in Australia [32,33], and so this sample may be biased toward representing those who are actively involved in cooking and grocery decision-making in Australia.
Other limitations include the self-reporting of vegetable intake, which is vulnerable to over-reporting [34], without prompts as to what to consider as a vegetable, which may conversely result in under-reporting. Classification as users and non-users based on selfreporting also lacks resolution, as frequency of use was not captured, and this may impact results. This requires further investigation. Frequency of recipe-following also requires more resolution in futher investigation to quantify potential differences in "sometimes" followers. However, strengths include the large sample size, the multiple measures of vegetable consumption (intake and variety), the multiple categories of convenience cooking product investigated, and the novelty of the questions, with potential implications for consumers, industry, and public health. The findings presented here justify further research into the role convenience cooking products play in diet quality.
Though convenience cooking products have the potential to help increase vegetable consumption in users, additional research is needed to provide a deeper insight into consumers' attitudes and actions towards vegetable consumption when using these products. A better understanding of usage habits and vegetable consumption drivers may assist the industry in adapting these products accordingly, to ensure an adequate vegetable intake and variety of vegetable intake. Overall, traditional methods of education and promotion have not improved population vegetable intakes at a population scale; as such, shifting paradigms are still needed to improve vegetable intakes, using products that meet consumer needs, and address common barriers to home cooking. The presented data, showing lower vegetable intakes with product use, with a higher number of product types used, and with more frequent recipe-following, but with little difference in related vegetable choice factors, provide an insight into the convenience cooking product users' vegetable intake, which provides a baseline for future improvements to the product back-of-pack recipe vegetable content in hopes to see an increase in the users' vegetable consumption.