Pilot Test of a Diet Cost Calculation Process for Use in Economic Evaluations of Dietary Interventions

ABSTRACT

Background

Nutrition is a significant modifiable risk factor for many illnesses. One often‐cited barrier to diet change is the perceived cost of healthy food. While many diet diary templates exist, none have been developed to capture the data needed for calculating the cost of a diet consumed during a diet intervention study.

Methods

The objective of the present project was to develop and pilot test (1) a tailored diet diary template with instructions, and (2) alternative approaches for calculating diet cost. We created a draft diet diary and instructions which were used by participants. We used those data to test the feasibility of calculating weekly diet cost. Two methods were used: calculation of the cost of the actual price of the items purchased, and standardised prices. Acceptability was assessed by conducting participant interviews and the feedback that was provided was used to revise them. The methods to calculate costs were compared in terms of feasibility and researcher time needed.

Results

Of the 15 participants enrolled, 14 completed the interview. All participants reported that the process was acceptable or highly acceptable. Participant feedback was used to guide changes to the diet diary and instructions. Costs were calculated for 85.3% to 96.3% of food items. The researcher time needed to complete each method per participant, per week was 5 to 6.5 h. Participants who gave feedback confirmed that the edits addressed their previous concerns.

Conclusions

Overall, the diet diary and instructions were acceptable to participants who are women and feasible for the collection and costing of diet data in Canada.

Summary

• This project involved the development and pilot testing of a process for calculating diet cost.

• Fourteen participants found the diet diary acceptable and provided feedback for revision.

• Calculation of diet cost was feasible using the data collected.

1. Introduction

Nutrition is a significant modifiable risk factor for the development and progression of many illnesses [1]; however, it is recognised that adherence to dietary recommendations can be poor [2]. While a range of barriers to diet change have been identified, one that is often cited by patients, health care providers, and in the literature, is the perceived cost of healthy food [3, 4, 5]. Studies on the cost of food have reported conflicting results with some reporting an estimated increased cost [6, 7, 8, 9] associated with healthy eating and others not reporting an increase [10, 11]. These inconsistent findings may be the result of different methods of assessing diet or calculating cost, different participant populations, different geographic locations, or the study of different diet patterns and recommendations [11]. Many studies have reported that higher food costs were attributed to foods that are promoted in dietary interventions (such as vegetables and fish) as well as foods that individuals may be encouraged to limit (such as red meat, full fat dairy, highly processed and restaurant foods, and sweets) [7]; thus, recommendations to improve overall diet quality may result in simultaneous increases and decreases which could impact total cost in different ways depending on the relative proportions of these foods. When studying the therapeutic effects of a specific diet, such as one delivered in a clinical trial, it is important for adherence to also consider the cost burden of the specific recommendations being given.

An abundance of research has been done to validate methods of diet data collection, including the diet diary [12]. The diet diary has recognised strengths, including the potential for a high level of detail, no requirement for an interviewer or observer, and minimal recall bias [13]. Because of the known day‐to‐day variability in diet, data from the diet diary can be more useful than from the 24‐h diet recall [10]. However, the diet diary also imposes a relatively high respondent burden, requires a longer period of time for data collection (usually 1 week), and creates the potential for participants to change their eating habits as a result of the process [13]. Despite significant research on the use of diet diaries for collection of nutritional data, there is a dearth of shared, tested resources or processes in the form of published templates and methods for their use in diet cost data collection and calculation for clinical trials of dietary counselling interventions. The general purpose of diet diaries is to help a clinician calculate nutrient intake and understand an individual's eating patterns. Factors such as whether produce is frozen, canned or fresh, or the brand or purchase location would not be relevant for a standard diet diary but are necessary for understanding diet cost.

Dietary interventions have been found useful in the treatment of various chronic illnesses, including cardiovascular disease, both in terms of therapeutic effectiveness and cost‐effectiveness [14]. However, authors of these studies have acknowledged limitations in their approaches, including use of incremental modelling or extrapolation of dietary intake from population references to identify the cost of diet or dietary change. Though findings position lifestyle approaches that include dietary components as strong cases for renewed focus, further research is needed, including in other health conditions and physiological roles for dietary interventions [15] and in ways that incorporate nuanced means of capturing elements of a dietary intervention. Dietary counselling is also emerging as a novel treatment approach for other common disorders where not all individuals respond to available treatment options. For example, not all individuals with depressive disorders or anxiety disorders, which are common [16], respond to the available treatment options [17]. Dietary counselling is emerging as a novel treatment approach for those populations. Randomized controlled trials have demonstrated that dietary counselling interventions based on the Mediterranean diet, alone or in combination with omega‐3 fatty acid supplementation, can significantly improve symptoms of depression in participants with depressive disorders [18, 19, 20, 21, 22].

Recently, our team completed the EASe‐GAD (Eating and Supplementation for the treatment of Generalized Anxiety Disorder) study, which was the first to assess the impact of dietary counselling plus omega‐3 supplementation on the symptoms of GAD [23, 24]. In addition to demonstrating the feasibility and acceptability of the intervention, a 55% reduction in anxiety symptom severity was observed [23]. In recruiting participants for the study, the EASe‐GAD team contacted local mental healthcare providers; several voiced the concern that food costs might prevent some individuals from being able to participate fully in the dietary counselling program that was being studied. While participating in the program, several participants spontaneously reported a reduction in grocery and eating costs. In qualitative interviews conducted as part of the pilot study [25], many participants commented that food cost was an important factor influencing peoples' ability to participate and that evidence of lowered food costs would make the intervention more attractive to future participants, particularly those of lower socioeconomic status. However, a formal assessment of food cost and cost‐effectiveness was not incorporated into the pilot study design which focused on testing the feasibility and acceptability of the intervention delivery.

In addition to considering the therapeutic effects of an intervention, the costs of the intervention must also be considered for at least two reasons. From a broad policy perspective, it is necessary to understand the cost of the intervention from a societal perspective – i.e., capturing all costs no matter who pays. The landmark nutritional psychiatry study, SMILES, reported a reduction in depression symptoms in response to dietary counselling [18] and found that average health sector costs were $856 lower and societal costs $2591 lower among patients receiving the diet intervention compared to participants in the comparison arm in their companion cost‐effectiveness analysis [26]. The cost difference in health sector costs was the result of fewer health professional visits. The cost difference from the societal perspective was due to the reduction in visits, lower costs related to unpaid productivity, and to lower food costs. The SMILES study's economic analysis included calculation of the cost of food prior to the intervention and an estimation of diet cost based on a hypothetical diet following the intervention recommendations [11]. They found that the cost of the recommended foods was less than baseline participant diet costs. However, because they did not calculate the cost of the participants' actual food during the intervention, it is difficult to draw definitive conclusions about the true cost impact [27]. Secondly, from the perspective of the participant, capturing food costs is important for the acceptability and affordability of the intervention, as well as adherence. Overall, there is a need for more economic evaluation of nutritional interventions to help inform and lower barriers for patients, and to support healthcare policy recommendations.

While analyses for assessing change in diet cost following an intervention typically use a standardised list of prices to value changes in resource use (here, changes in the amounts and type of food purchased), we recognised that recommendations related to purchasing food may be included in a dietary counselling intervention, as they were in the EASe‐GAD study. As a result, participants may change both the types of foods that they eat, as well as how, when and where they purchase those foods. In the EASe‐GAD study, participants received a handout titled “Healthy Eating on a Budget” [28] which included recommendations about purchasing foods such as making a grocery list, purchasing items on sale, purchasing canned items, comparing prices, shopping at discount grocery stores, and choosing foods that are in season. Because the intervention included these recommendations, a cost approach using standardised pricing may not accurately capture the true change in food cost that occurred as a result of the intervention. Conversely, there may be benefits to assessing diet cost using standardised methods, such as 1) separating the cost impact of the altered purchasing habits from the change in food itself, 2) decreasing participant burden if participants are not required to record purchase information, and 3) offering a mechanism to calculate a price in the event that insufficient information is available to calculate the actual price paid. Thus, consideration of multiple methods of cost calculation is warranted. The objective of the present project was to develop and pilot test 1) a tailored diet diary template with instructions, and 2) approaches for calculating diet cost, which can be used in clinical trials to measure the difference in cost between baseline and intervention diets. To do this, we assessed the feasibility of using our draft template to calculate weekly food cost, assessed its acceptability to participants, and refined the design based on participant feedback. We intend to use the final version to calculate the cost of a diet consumed by study participants prior to and during a subsequent EASe‐GAD dietary counselling intervention study, but also offer it to other researchers to use in dietary intervention studies.

2. Methods

This project included the design and pilot testing of a process for diet data collection and cost calculation. This project was overseen by the Research Ethics Board of the Canadian College of Naturopathic Medicine (CCNMREB059.Aucoin). Informed consent was obtained prior to participation.

2.1. Initial Diet Diary Design

A diet diary template with instructions for participants were drafted by the interprofessional study team, led by a registered dietitian (RDN). The design was based on standard practice in dietetics and a clinically validated diet diary template [29, 30]. The diary included space to record standard diet diary information such as the food eaten and the portion size consumed, but also included space to record information specific to allowing cost calculation, such identifying canned or frozen ingredients, which might differ in price from fresh ingredients, and the brand of the food item and the location of purchase. The instructions provided directions for completing the diary for seven consecutive days. The study team beta‐tested the diary and instructions for face validity, and through discussion, reached consensus on the components and organization of the template and instructions. (See Supplemental File 1 for the draft version of the diet diary and instructions).

2.2. Diet Diary Pilot Testing

In the pilot testing phase, participants were recruited and provided with the diet diary and instructions. We aimed to recruit 15 participants. They were asked to complete it for 7 consecutive days and return the document by email. Participants who did not return the document within 2 weeks received up to two follow‐up attempts. All participants, regardless of whether or not they had returned the diet diary, were invited to attend an individual, virtual, 30‐to‐60‐min interview. It was conducted by a member of the research team using a semi‐structured interview guide developed for this project. The participants were be queried about:

• Their experience completing the diary

• The number of minutes spent during each session of recording data and if they felt that it was reasonable

• The number of times per day that they recorded data

• Meals or food items that they found difficult to record on the template

• Components of the template or instructions that were unclear

• Opportunities for improvement of the template or instructions

• They were asked to rate the process of recording diet data as highly acceptable, acceptable, unacceptable or highly unacceptable, and to provide a reason for this response.

Following revision of the diet diary and instructions, we invited participants who had expressed suggestions for improvement to do a brief revision checking interview. This involved reviewing the revised diet diary and instructions and a 15‐min interview asking for feedback on the revisions.

2.3. Participants and Recruitment

All eligible participants from the previous EASe‐GAD study [23, 24] were contacted and invited to participate in testing of the diet diary. The first 15 individuals who expressed interest in participating were invited to complete the informed consent process. Because we had previously collected demographic data, including socioeconomic status, from all prospective participants, we planned to recruit a minimum of three participants from each quartile of income.

2.4. Data Extraction and Cost Calculation

The data from the returned diet diaries were extracted into an Excel spreadsheet where we identified a price (unit cost) for each food item that had been consumed using each of the two methods described below, applied those unit costs to the amounts consumed of each food item, and totaled all item costs to get a weekly food cost. To assist the pricing process, the portion size reported for each item was converted to a format that was consistent with how that food item is packaged and sold (e.g. per ml, per gram, per item, etc). Unit cost data was obtained, using the methods described below, and the unit cost of each food item was multiplied by the serving size consumed (see Supplemental File 2 for the Excel template and example and Supplemental File 3 for instructions for using the Excel template).

The a priori plan for assessing the weekly diet cost was to calculate costs using two methods ‐ the first method involved calculating the actual cost of the food items purchased, while the second used standardised costs for each food item.

2.4.1. Method 1

The cost of each individual item was obtained from identified grocery store or restaurant websites or databases based on the purchase location reported by the participant. When pricing information for the exact item is not available, a comparable store's pricing or market basket database pricing was utilised. Price data were obtained from stores and restaurants located in the geographic region of the pilot study. When multiple packaging sizes were available, we used the following approach which is similar to that used in previous research studies [9]. When considering products with a short shelf life, such as eggs or milk, cost data for small package sizes were used. For food items with a longer shelf life, such as dried grains or olive oil, cost data for a larger package size were used. For foods with a moderate shelf life, a mid‐range package size was selected.

2.4.2. Method 2

A standardised list of prices was used for each item. That price was used whenever the food item was consumed by any participant, regardless of the brand or place of purchase recorded by the participant. We used Statistics Canada's Monthly Average Retail Prices for Selected Products [31] to create a standardised price for food items. This method provides a reliable and standardised source for estimating food item prices consumed, as it reflects actual consumer retail prices collected across Canada using consistent methodology.

As we applied this costing algorithm, we identified, developed and documented reasonable assumptions. In cases where participants frequently shopped at a single store and there was a store left blank for a food item, the frequently reported store was used. If cost data for the exact item could not be found at the store that was reported, price data from a similar store was used (i.e., we categorised local stores as premium, high‐end, mid‐range, discount, and warehouse). Sale prices were not used as these varied between the time of purchase and time of cost calculation. Tax was included if applicable for the item. Surcharges for grocery delivery were excluded.

2.5. Feasibility

To assess feasibility, we assessed the percentage of items for which we could calculate a cost using the above‐described methods. We considered the method feasible if we could calculate the cost for at least 80% of food/beverage items. Areas where insufficient data were available for pricing were highlighted in order to inform revision of the diet diary and instructions. We also compared the time required to complete the two cost methods.

2.6. Acceptability

The data from the participant interviews were used to assess acceptability. The interpretive framework used in the qualitative analysis was pragmatism, as the main goal of the study was to capture data that can be used to improve food cost capture in subsequent follow up studies. Participant feedback from the interviews was analysed using Structured Framework Analysis [32] to inform diet diary and instruction revisions. This approach is commonly used in evaluation studies. For each element of the diet diary and instructions (e.g., format, instructions, burden) we identified positive and negative response data and suggestions for improvement. The question about satisfaction was used to assess acceptability. The cutoff for acceptability was set at 80% of participants stating that the process was acceptable or highly acceptable.

3. Results

Seventeen individuals expressed interest and the first 15, which included three participants from each quartile of income, were enrolled. Twelve participants completed and returned their diet diaries (Figure 1). Fourteen participants completed an interview. Data were analysed from all interviews. The cost calculations were completed for 5 diet diaries as it was deemed that the process of analysing the diet diaries was highly similar between participants. Three participants completed the revision checking interview.

Figure 1.

Flow of participants through the study.

The demographic characteristics of the study participants are presented in Table 1. The mean age was 44 years (SD 12) with the participants being most frequently married (47%), employed (73%), university‐educated (73%), and white (40%). Overall, the participants in this study were similar to those who participated in the EASe‐GAD [23].

Table 1.

Demographic characteristics of study participants.

Demographics	Number (%)
Mean Age (SD)	44 (12)
Sex‐assigned at Birth
Female	15 (100)
Marital Status
Single	5 (33)
Married/Common‐law	7 (47)
Divorced	3 (20)
Employment Status
Employed full‐time	11 (73)
Employed part‐time	2 (13)
Unemployed	1 (7)
Retired	1 (7)
Ethnicity
Black (African, Afro‐Caribbean, African Canadian descent)	2 (13)
East/Southeast Asian (Chinese, Korean, Japanese, Taiwanese, Filipino, Vietnamese, Cambodian, Thai, Indonesian descent), Latino (Latin American, Hispanic descent) Middle Eastern (Arab, Persian, Afghan, Egyptian, Iranian, Lebanese, Turkish, Kurdish descent), or Indigenous (First Nations, Metis, Inuk/Inuit descent)a	3 (21)
South Asian (East Indian, Pakistani, Bangladeshi, Sri Lankan, Indo‐Caribbean descent)	2 (13)
White (European descent)	6 (40)
Another race category	2 (13)
Education
No certificate, diploma or degree	0 (0)
High school diploma or equivalency	0 (0)
Apprenticeship or trades certificate or diploma	0 (0)
College or other non‐university certificate or diploma	4 (27)
University diploma or certificate below bachelor level	0 (0)
University certificate, diploma or degree at bachelor level or above	11 (73)
Income
$10,000 or less	0 (0)
$10,001 to $20,000	3 (20)
$20,001 to $30,000	0 (0)
$30,001 to $40,000	0 (0)
$40,001 to $50,000	0 (0)
$50,001 to $60,000	1 (7)
$60,001 to $70,000	0 (0)
$70,001 to $80,000	3 (20)
$80,001 to $90,000	3 (20)
$90,001 to $100,000	1 (7)
Greater than $100,000	4 (27)

^a To protect participant confidentiality, categories with small cell size (zero or one participant) were aggregated.

3.1. Acceptability

Overall, all participants were satisfied with the diet diary, with 58% reporting its use as “highly acceptable” and the remaining 42% reporting its use as “acceptable”. Table 2 presents the qualitative data from the participant interviews.

Table 2.

Qualitative data from participant interviews.

	Positive	Negative	Suggestions for improvement
Diary Design Format	‐acceptable layout	‐difficulty accessing without Microsoft Word	‐create a version that doesn't require MS Word
	‐easy to use	‐hard to add additional lines	‐make it accessible on a mobile device/phone
	‐easy to navigate		‐include instructions on how to add additional lines
Information Gathered	‐reasonable level of detail	‐hard to report brand if packaging had been discarded	‐opportunity to record amount paid for items purchased during the study period
		‐surprised that we didn't ask about price paid
Instructions	‐clear instructions	‐uncertain about level of detail required	‐improve instruction for reporting home‐made food (especially when cooking a batch) and restaurant items
	‐examples and images were helpful		‐clarify if portion size is cooked or uncooked
			‐include a brief training session to review the instructions with participants
			‐add more examples to the instructions to illustrate level of detail
Participant burden	‐acceptable amount of time	‐required motivation	‐method for repeating meals rather than re‐recording
		‐repetitive when recording the same meal multiple times
Barriers	n/a	‐left the diary at home	‐make it easier to complete on a mobile device
		‐did not turn on computer on the weekend
		‐forgetting
Facilitators	‐making notes on phone and then transferring to the diary at a later time	n/a	‐include instruction about taking photos if not able to record immediately
	‐took photos of food with phone
Other	‐increased awareness/mindfulness of eating habits	‐some reported an impact of food choices	‐include instructions letting participants know about the possibility of these experiences
		‐concern about judgement from those reading it

When participants were asked about the amount of time spent completing the diet diary, the most common frequency of recording diet information in the diet diary was once daily and the most common amount of time reported was 15–30 min.

Many participants provided positive comments about the diet diary and instructions. “I thought everything was helpful. I thought everything was very clear. It was not hard to navigate at all” (participant 23). “Yes, I liked the portion size and how it had the visuals” (participant 48).

Many participants highlighted areas that were unclear, “I wasn't sure how much detail you guys wanted.” (participant 48) “I didn't know how much a detail was very important to you, so things like the salad dressing that I make on my own. I was sure I had the right amounts. And then, I think, on the sidebar I recorded there was left over because I didn't want that that volume was consumed on that salad. So that was the thing that was tricky. Because I had to kind of record what it was to make it. And then it was kind of like. No, no, but there was left over.” (participant 23) “the prepared items were a bit difficult because sometimes it's like cultural food that has a lot of separate ingredients” (participant 48). “The only part for me where I had a little bit of difficulty was when it was the bigger batches and putting it into smaller portions and I guess maybe like I didn't have enough space.” (participant 49) “If there were meals that had a little more preparation or recipes that had more ingredients or spices. That was a little bit more challenging. But I wrote it up, and it wasn't that difficult, actually.” (participant 24) Other comments demonstrated that participants had not understood the instructions. Despite instructions that restaurant meals did not need to include ingredients, several participants commented reflected a lack of clarity “for the meal that I got at the Food Court. I don't know where they got their bean sprouts, or their onions, or tomatoes, or the garlic, or the oil that they use, or the spices.” (participant 50)

Participants spoke to the level of burden and ways to increase convenience, “So for me, having a busy lifestyle, it's like a little bit time consuming but interesting.” (participant 29) “I'm always on the road. So for me, it's just easier to like do it on my cell phone right away. I know, like an app is probably too much to ask for a study research project. But if it was just in a way that was very simple, so that you can just fill it in easily on your cell phone.” (participant 48) “I think that the only reason why I'm not going highly acceptable is just because everyone is on their phone. And we're so used to recording it on the phone. This, not the struggle, but the inconvenience. You can call it to open the laptop, open that word document and record it and save it just made it a little bit more difficult. And then, you know, sometimes you forget, I was saying Saturday and Sunday. I didn't work. I didn't open my laptop. So I'm not gonna open just for the sake of recording my food habits. So I thought, I'll do a catch up on Monday, and I just have to remember.” (participant 5)

Although it was not the purpose of the diet diary, many participants reported increased awareness of their eating habits. Many felt that this was a positive feature of participating in the project. “it makes you more mindful of eating healthy. I found it kind of helped me. I look back and I thought, oh, that was good. I did home some home cooking, and sometimes, you know, I went to [fast food restaurant], but it was overall good” (participant 24). “I think it's highly acceptable, because I think it helps you to reflect. I think when words are written it's a better understanding of what I'm consuming and what I'm doing. I can reflect back and see Wow! That was healthy, or that wasn't healthy, or that was, you know, prepared with wholesome ingredients. Or it wasn't” (participant 24) “Yeah, it's highly acceptable that way, because you could use it as a learning tool to how you spend your money and your eating habits. And yeah, I think that would help” (participant 49).

3.2. Feasibility of Data Extraction and Cost Calculation

Overall, the diet dairies had a high level of completion; 95% of text fields were complete on the returned diaries.

Per participant, it took researchers approximately 2.5 h to enter 1 week's worth of diet data into the Excel table and convert the portion size to the most commonly used units. The amount of time required to complete the cost analysis varied by method. It took longer to perform cost calculations using Method 1 than using Method 2. The time required to access item prices from individual websites was the primary factor. The amount of researcher time required to complete the cost calculation after data entry, per‐participant, was 4 h using Method 1 and 2.5 h using Method 2.

In attempting to complete cost calculations for five participants using Method 1 (i.e., using the brand and store of purchase recorded), costs were able to be generated for 85.3% of food items entered. In addition to the large amount of time required for this method, another challenge was finding prices for local or independent stores or restaurants that did not publish prices on a website. Using Method 2 (i.e., standardised database pricing for each food item) we were able to locate a price for 96.3% of food items. Comparison of the total weekly food price between the two methods resulted in a difference of less than 5% in food price.

3.3. Dietary Diary and Instruction Modification

Based on the data obtained from the participant interviews and the lessons learned in the process of calculating cost data, several revisions were made to the template, instructions and process (Supplemental File 4). To improve accessibility and ease of use, we created two versions of the diet diary including both a MS Word document and a GoogleDocs version so that participants who do not have access to MS Word would not face a barrier to access. The GoogleDocs version was tested for functionality on desktop and mobile electronic devices. We modified the instructions (found at the beginning of the diet diary template) to improve clarity about how to record restaurant foods and batch preparation of meals. We added instructions on how to repeat meals to decrease redundancy and participant burden and revised the example to include some meals that participants reported were more challenging to record. We also revised the instructions to increase clarity about measurement. We added instructions for participants on how to include price data from purchases that they made during the study period, although noting that this was optional. Lastly, because it was clear in the interview process that some participants had not read the instructions at the beginning of the diet diary, we generated a script for researchers to read to participants to provide some brief orientation and increase the likelihood of the participants correctly following the instructions. (Supplemental File 5).

In the revision checking interviews, participants reported that the revisions improved the usability and clarity of the diet diary and instructions. They identified a small number of formatting suggestions and indicated that they were satisfied with the revisions.

4. Discussion

The process of using the diet diary and instructions was considered acceptable to the study participants and their feedback resulted in meaningful refinement of the instructions and diet diary. The data generated were also feasible for calculating weekly diet cost.

The feedback from participants resulted in meaningful changes to the diet diary and instructions. Some areas of uncertainty by the participants reflected opportunities to improve the clarity of the diet diary and instructions, while others highlighted the fact that some participants had not carefully read the instructions. As such, we created a script that could be read to participants as a brief orientation prior to completing the diet diary to ensure that all participants were aware of the instructions and had the opportunity to ask questions before beginning. Additionally, the creation of a web‐based version helped to decrease barriers to accessibility.

Calculating diet cost using each of the two methods was considered to be feasible, as we were able to find a cost for more food items (85.3% and 96.3%) than our pre‐specified target (80%). Each of the two methods has both strengths and limitations, and the correct approach to be used in a subsequent study will depend on the research question being asked. The standardised approach (Method 2) was faster to analyse, however, it does not account for changes in participant purchasing behaviours. Although the individualised approach (Method 1) requires a larger amount of time to analyse, it allows the researcher to assess the impact of an intervention that makes recommendations about the type of food, as well as how or where to purchase it. The cost of food is commonly cited as a barrier to adopting healthier eating patterns. To best answer the question of whether it is possible to improve the quality of a diet without increasing cost, the individualised method may be necessary. The small difference (< 5%) in diet cost between the two methods indicates that our source of standardised prices provides a reasonable estimation of actual prices, and that these standardised prices may be a reasonable substitute for missing actual prices when using Method 1.

One strength of this project was the intentional inclusion of participants from diverse socioeconomic backgrounds. This helped to identify economic barriers to access (i.e. availability of Microsoft Word application) within the original version which were addressed in the revisions. Secondly, the project was completed by a multidisciplinary team who provided diverse perspectives and experiences. The project involved participation of a clinical population. The high level of participation in the interview suggests that the high level of satisfaction is not due to selection bias.

This project had limitations. Given the utilisation of a standard diet diary as a starting point, the project is limited by the established challenges inherent in diet diaries including participant burden and increased self‐awareness [13]. While the modifications suggested by the participants were very clear and reviews of our revisions by the participants were positive, we did not repeat our complete testing process with the revised version. Also, the sample size participating in all data collection was relatively small. The participants in this project had previously participated in the EASe‐GAD study which was limited to adult women with moderate to severe GAD. It is possible that participants of other ages or genders or with other medical conditions may have different needs and views about the diet diary developed, particularly given the gender roles that have historically existed related to household food preparation. Researchers planning to use these materials with different patient populations may consider consultation with a representative of the patient population to identify if any adaptations are warranted. Additionally, researchers in other geographic locations may have different access to food pricing data for their local region, impacting feasibility. While Method 1 is able to capture changes in food cost related to purchasing food at different stores, it does not capture efforts by the participant to purchase sale items. The cost calculation process involved a relatively high burden on researcher time for obtaining and analysing cost data; however, future research could consider ways to incorporate automation, use of technology or integration with other digital platforms. Lastly, this work involved assessment of feasibility and acceptability; further formal evaluation could be undertaken to test validity and cost calculation accuracy.

5. Conclusions

The developed diet diary and instructions were feasible for collecting data for cost calculation in Canada and acceptable to women participants. The participant feedback resulted in meaningful modifications to instructions and format. The developed resources can be used for calculating weekly diet cost to allow comparisons of the cost of baseline and intervention diets. This will enable researchers to measure the food cost implications associated with dietary interventions being studied while also capturing data relevant to dietary analysis.

Author Contributions

Monique Aucoin: conceptualization, methodology, formal analysis, writing – original draft. Hainan Yu: methodology, formal analysis, writing – original draft. Melissa Murphy: methodology, formal analysis, writing – review and editing. Nicole Yoannou: formal analysis, investigation, writing – original draft. Laura LaChance: formal analysis, writing – review and editing. Kieran Cooley: conceptualization, methodology, formal analysis, writing – review and editing. Patricia M. Herman: methodology, writing – review and editing.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File 1

Supporting File 2

Supporting File 3

Supporting File 4

Supporting File 5

Aucoin M., Yu H., Murphy M., et al., “Pilot Test of a Diet Cost Calculation Process for Use in Economic Evaluations of Dietary Interventions,” Journal of Human Nutrition and Dietetics 39 (2026): e70302, 10.1111/jhn.70302.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.