Unhealthy diets contribute to increasing obesity and diet-related non-communicable diseases (NCDs) [1]. The cost of food is a major determinant of food choices [2, 3]. Some countries have implemented health-related taxes or subsidies in an effort to improve population diets [4]. Taxes on sugar-sweetened beverages (SSB) are increasingly common internationally [5], and have shown to significantly reduce SSB purchases, especially in lower socio-economic population groups [6, 7]. In addition, some countries do not tax healthy foods, for example, in Australia, there is no Goods and Services Tax (GST) on basic healthy foods [8].

Evidence on whether or not healthier diets or dietary patterns are more expensive than less healthy diets is mixed [9]. Currently there is no robust methodology available to adequately answer that question. Most previous studies did not sufficiently take into account the variation in possible diets or food prices, or measured only the cost of the healthy diet [9]. Healthy diets are usually based on food-based dietary guidelines, and are developed by substituting items in a typical diet with healthier items, by developing a diet to meet food-based dietary guidelines or based on a typical diet of those who meet dietary guidelines. Some studies used linear programming to develop healthy diets that meet nutrition recommendations for a minimum cost [10]. There are a few previous studies that did measure the cost differential between healthy and current diets, but they compared the cost of only one healthy diet with one current, less healthy population diet [11–15]. The variation of the cost of diets is important when considering the relative cost differential between healthy and current diets, but is currently unknown. Many diet scenarios can be constructed using a list of commonly consumed foods to meet nutrient and food-based dietary guidelines (for ‘healthy’ diets) or specified population nutrient and food intakes (for ‘current’ diets). The International Network for Food and Obesity/NCDs Research, Monitoring and Action Support (INFORMAS) [16] developed a useful framework to monitor the cost differential between healthy and current population diets globally [17]. Such monitoring aims to provide robust data and benchmarks to inform economic and fiscal policy responses.

Within the INFORMAS framework for monitoring the cost of population diets, this study developed a new tool and methodology, DIETCOST, and modelled, for the first time, the cost differential between current household diets and healthy household diets in New Zealand, taking into account the variations in two-weekly household meal plans and food prices. The cost differential between healthy and current household diets was compared between different ethnic population groups and for a series of diet and prices scenarios. In addition, the potential impact of two policy scenarios (fresh and frozen vegetables exempt from GST and a tax on SSBs) on the cost differential was modelled, and the Food Price Index (FPI) data was used to model the cost of current New Zealand household diets over a ten year period.

The study was approved by the Human Participants Ethics Committee of the University of Auckland (ref 12330). A novel DIETCOST programme [18] was developed for researchers, using Python, to model the cost of healthy and current household diets using a list of commonly consumed foods, a set of min and max quantity/serves constraints for each, and specified food group and nutrient intakes based on dietary guidelines (healthy diets) and nutrition survey data (current diets for different population groups).

The programme was applied in New Zealand as a case study for the total population and specific ethnic population groups.

The reference household comprised a 45-year old man, a 45-year old woman, a 14-year old boy and a 7-year old girl.

Interface

The programme user interface [18] allows the user to specify the daily targets for the food groups and nutrients for all household members for current and healthy diets separately. In addition, the interface allows the user to specify whether or not to include takeaway meals, alcohol and discretionary foods as part of the diets.

The minimum serve size difference between any two generated individual meal plans was set at half a serve for any common food in this study.

The programme algorithm uses the Mersenne Twister as a random number generator to specify the starting meal plan and the starting value in grams for each of the common foods. If a meal plan meets all targets/constraints and is not already in the list of matching meal plans, it is added to the results. If it doesn’t (i.e. it fails some constraint), the algorithm will then try to fix that constraint (by raising/lowering the amount of some item that affects that constraint randomly between the min and max amount for that food item). If the modification results in a matching meal plan the meal plan is added to the results, and the same procedure starts again until the specified number of iterations has been run (Fig. 1). If the modification does not result in a matching meal plan, the algorithm will continue to try to resolve one of the failing constraints in a subsequent iteration. All success meal plans are independent from each other.

For each individual household member, the current and healthy diet scenarios were run multiple times with 1million, 2million and 20million iterations respectively to find the right number of iterations needed to accurately estimate the average cost of healthy and current household diets.

About 1million iterations allowed an accurate estimation of the average cost of healthy household diets, while about 2million iterations were needed for an accurate estimation of the average cost of current household diets in New Zealand.

Standard healthy and current diets for the total New Zealand population

The energy density of the average current household diet was about 50% higher than for the average healthy household diet. Current diets were all found less healthy than the healthy diets as, for example for the adult men, 0 meal plans met all healthy diet guidelines due to 0 meal plans meeting maximum sodium intake, minimum fibre intake and minimum number of serves of fruit (Additional file 1: Table S1).

The average fortnightly cost of healthy New Zealand household diets was about $27 more expensive than the average cost of current New Zealand household diets (Table 1). However, about 25.8% of healthy diets were cheaper than the cost of the average current household diets, and the cheapest healthy household diet was cheaper than the cheapest current diet (Fig. 2). The variation around the average cost was greater for healthy than for current diets and the range of costs of current diets fell within the range of costs of healthy diets (Table 1; Fig. 2).

	Total NZ population		Pacific population		Māori population
	Healthy diets	Current diets	Healthy diets	Current diets	Healthy diets	Current diets
N iterations	1,000,000	2,000,000	1,000,000	2,000,000	1,000,000	2,000,000
N common foods included	73	100	71	105	75	109
N individual meal plans	365 + 347 + 319 + 137	360 + 185 + 96 + 153	120 + 226 + 299 + 281	252 + 452 + 330 + 329	93 + 137 + 17 + 44	332 + 127 + 12 + 367
N household meal plans	5,960,900,400	978,220,800	2,278,595,280	12,366,557,280	9,530,268	185,690,256
Average cost (SD)	$723.4 (75.7)	$696.3 (48.5)	$593.7 (65.0)	$655.1 (62.6)	$655.3 (65.0)	$693.6 (43.1)
Range of the cost	$502.0 - $937.1	$571.1 - $850.0	$423.4 - $778.3	$448.6 - $799.6	$522.0 - $790.7	$580.3 - $796.3
Energy (MJ)b	139.6 (137.6-141.7)	152.2 (150.3-154.8)	139.6 (137.6-141.7)	165.7 (163.8-168.7)	139.7 (137.6-141.4)	161.4 (159.4-164.1)
Energy density (kJ/g)b	1.02	1.49	1.03	1.56	0.94	1.65

SD standard deviation

^aNo alcohol, takeaways or discretionary foods included in the standard healthy diet; takeaways and discretionary foods included in the current diet

^bAverage (range)

Discretionary foods (including beverages) and takeaways contributed about 35.5% to the cost of the average current diet, while fruits and vegetables contributed about 19% to the average cost. Healthy diets did not contain takeaways and discretionary foods but fruits and vegetables contributed about 40% to the average cost of healthy diets. Protein sources were the major contributor to the cost of healthy and current diets and contributed on average more or less one third to the cost of both healthy and current diets, while dairy contributed about 5% to the cost of current versus 13% to the cost of healthy diets (Fig. 3).

Trends in the cost of the New Zealand population diet over time

Using the Food Price Index data over 10 years, the cost of current diets was a bit more expensive, since not all prices collected as part of the FPI are necessarily the cheapest prices. Based on this data, the cost of current household diets substantially increased over time. The contribution to the cost of different food groups was very similar across seasons though vegetables tend to contribute more to the cost in winter (Fig. 4, Additional file 1: Fig. S1).

This study developed a new tool and methodology, DIETCOST, to model the cost differential between healthy and current population diets, and applied it as a case study in New Zealand.

For the New Zealand population, healthy diets were on average $27 more expensive than current diets over a fortnight in New Zealand, but one-quarter of healthy diets were cheaper than the average current diet. For Māori and Pacific population groups, however, in view of their high current energy intakes, current household diets were on average $40-60 more expensive than healthy diets. Healthy diets for Māori and Pacific households are generally cheaper than for the total population because of several factors, e.g. prices in South Auckland supermarkets and fresh produce markets were cheaper (where Pacific populations live and shop), some fruits and vegetables, like feijoas, puha, kamo kamo, were included in the Māori diets for free since they would be gifted or gather these foods, and healthy Māori diets did not contain lower salt/fat/sugar versions of products.

Strengths of the study include the development and validation of a new programme, DIETCOST, to generate shopping lists for two-weekly meal plans that meet the targets and constraints for both the current and the healthy diets. In addition, common foods and nutrient and food group targets for different ethnic groups allowed conducting the modelling for specific ethnic populations. Unlike studies to date that have compared the cost of one healthy and one current diet, DIETCOST allows the cost of many fortnightly household diets to be generated enabling comparing the distribution of costs of current and healthy diets. DIETCOST provides a tool to calculate the cost of many meal plans for a range of scenarios of changing the diet contents, altering the type of price, location or reference household, without the need to do this manually. The programme can be readily used and applied with other population groups in other countries and contexts after adapting the input files. Limitations of the study include the fact that New Zealand lacks up-to-date nutrition survey data as the latest survey was conducted in 2008. A new nutrition survey needs to be conducted urgently and was a key recommendation in the latest Healthy Food Environment Policy Index report [36]. The Food Price Index is a great resource to look at trends over time, but includes some clear limitations, such as the fact that some key foods are missing and it is not always easy to divide the foods into healthy and less healthy due to insufficient details in the description.

This study adds to the literature through providing a common tool to assess the cost differential between current and healthy diets. Previous studies only priced one healthy and one current diet and used different approaches to identify the diet, define healthiness and calculate the cost [9].

In Australia, it was found that the current diet was more expensive than the healthy diet with more than half of the current household budget for food spent on energy-dense, nutrient-poor foods, which is more than in New Zealand [11]. It needs to be noted however that in Australia fruit and vegetables and basic healthy foods are exempt from GST. In addition, there was no adjustment for underreporting, while this study calculated energy intake based on the current weight of the population rather than using survey data. In Denmark, it was found that the healthy New Nordic diet was about 16% more expensive than the current diet with the largest relative difference for low-income households [37].

The cost of food is an important determinant of food choices. This study developed a novel methodology to model the cost differential between healthy and current household diets. Healthy New Zealand diets are on average more expensive than current diets, but not for specific ethnic population groups.

Reducing taxes on fresh and frozen fruit and vegetables in New Zealand in conjunction with a sugary drinks tax could reduce the average cost of healthy diets towards the average of current diets and make it easier for people to consume healthy diets. The same approach as in this study can be applied in countries globally to benchmark the cost differential between current and healthy diets among countries internationally.