Diabetes is a metabolic disease characterised by hyperglycaemia due to decreased insulin secretion, insulin action or both(Reference Evert, Boucher and Cypress1). Data published by the WHO support that diabetes was the seventh leading cause of death in 2016(2). Moreover, recent findings demonstrated that in 2017, 451 million people between 18 and 99 years of age were diagnosed with diabetes worldwide, and these numbers are expected to increase and reach 693 million by 2045(Reference Cho, Shaw and Karuranga3). In Cyprus, in 2017, there were 93 200 cases of diabetes among the total population of 884 480. Furthermore, the prevalence of diabetes among the Cypriot population equals 10·5 %(4).
Diabetes and its complications pose a great financial burden for patients, families and society(Reference Danaei, Lawes and Vander Hoorn5–Reference Wu, Wen and Qin7). A recent study indicated that, in 2017, the total costs for healthcare of people with diabetes were extremely high, equalling $850 billion(Reference Cho, Shaw and Karuranga3). A medical nutrition therapy (MNT) is a nutrition-based treatment provided by a registered dietitian (RD), including nutrition diagnosis as well as therapeutic and counselling services, to help manage diabetes(Reference Morris and Wylie-Rosett8). The American Diabetes Association (ADA) supports that MNT can result in cost savings for the national health system and improved clinical outcomes(Reference Evert, Boucher and Cypress1). Particularly, a previous study has shown that MNT had a significant cost-efficient advantage and provided significant clinical improvements compared with the basic nutritional therapy(Reference Franz, Splett and Monk9). Although MNT is considered a keystone for the medical treatment of diabetes, relative literature is scarce.
Food insecurity exists when people have limited or uncertain availability of nutritionally adequate and safe foods and/or low affordability of foods fulfilling basic energy requirements(Reference Bickel, Nord and Price10). Most at risk of experiencing food insecurity are low-income groups and other social groups associated with poverty, such as single mothers(Reference Coleman, Rabbitt and Gregory11,Reference Tarasuk, Mitchell and Dachner12) . The risk of experiencing food insecurity among people suffering from non-communicable diseases (NCD), such as diabetes, has not yet been examined, although the prevalence of food insecurity and low affordability has been extensively and deeply examined among healthy populations of different nations(Reference Friel, Walsh and McCarthy13–Reference Chrysostomou and Andreou18). Recently, a study by Chrysostomou et al. has shown that the Cypriot Gluten-Free Healthy Food Basket (GFHFB) is costly and not affordable among low-income Cypriots diagnosed with celiac disease, and thus, patients are likely to experience food insecurity, compromising their long-term health(Reference Chrysostomou, Andreou and Andreou19).
A food basket, which is a mixture of basic products in sufficient amounts to adequately fulfil the energy requirements of each member of the family, is one of the most commonly used tools to monitor trends in the affordability of foods(Reference Goedemé, Storms and Van den Bosch20,Reference Goedemé, Storms and Penne21) . Based on the literature, no study has evaluated the cost of a realistic monthly food basket specifically developed for people with diabetes. Thus, the main objective of this study was to assess the cost, acceptability and affordability of the Cypriot Diabetic Healthy Food Basket (DHFB) and to examine whether the low-income diabetic population in Cyprus are at risk of experiencing food insecurity due to low affordability.
Methods
The aim of this study was to develop a DHFB on the basis of HFB developed for Cyprus by the same researchers(Reference Chrysostomou and Andreou18). The Cypriot HFB was constructed by a RD based on the National Guidelines for Nutrition and Exercise (NGNE) developed by the Nutrition Committee of Cyprus(22). NGNE were developed based on current scientific evidence, such as the results of existing epidemiological studies in Cyprus, the US dietary reference intakes (DRI) and the WHO/FAO nutritional guidelines. The acceptability and feasibility of Cypriot HFB was tested through focus groups (FG). An FG consisted of people of different socioeconomic and educational status. After consultation with FG, changes in HFB were done in collaboration with the RD as long as they did not contradict healthy eating recommendations.
Development of the Diabetic Healthy Food Basket
In the current study, baskets (DHFB) for two different household types were developed: those with one adult woman (±40 years) or those with one adult man (±40 years). It was assumed that both adults (woman and man) were diagnosed with diabetes. In particular, the development of food baskets should be based on specific nutritional guidelines. Therefore, DHFB was based on Cypriot HFB with adjustments based on the most recent nutritional guidelines for diabetes as developed by the ADA(Reference Evert, Boucher and Cypress1). ADA guidelines published in 2014 are the most recent for diabetes mellitus globally. Thereafter, the acceptability of DHFB was examined through the questionnaires. In comparison with Chrysostomou and Andreou (2016)(Reference Chrysostomou and Andreou18) and Chrysostomou et al. (2017)(Reference Chrysostomou, Andreou and Polycarpou23), the acceptability of our DHFB was examined through questionnaires. The advantage of this approach is the collection of information from a sufficiently large and representative sample of diabetic population in the country of reference. Based on information collected through the questionnaires, any changes in food baskets could be made only if ≥50 % of the participants supported the change. The change could be made in collaboration with the RD as long as it did not contradict eating recommendations for diabetes(Reference Chrysostomou and Andreou18).
Sample recruitment and data collection
Participants were recruited from public hospitals in Cyprus (Nicosia General Hospital, Limassol General Hospital) during 2018–2019 by the RD. All participants included in the study were aged between 18 and 87 years. To be eligible, they had to be diagnosed with diabetes mellitus (type 1 or type 2) regardless of treatment method and year of diagnosis. In total, 422 adult women (41·1 %) and men (58·6 %) signed the informed consent, which has been approved by the Cypriot Bioethics Committee and the Office of the Commissioner for Personal Data Protection.
A questionnaire inquiring on general social/demographic/financial, medical and nutritional aspects was administered. The questionnaire was based on FG discussions in a previous work(Reference Chrysostomou and Andreou18). The questionnaire included questions relating to the socioeconomic and demographic background of participants; other questions related to participant’s medical status such as the type of diabetes, age of diagnosis, HbA1C level, type of medical treatment, presence of other diseases, health problems among other first-degree family members, medical insurance, personal aspects about previous/current/future health status, complications of medical treatment and the effect of disease on financial status. In the third part of the questionnaire, questions relating to the DHFB were included, such as the adequacy of the food basket, food items to be added/deleted, taste/variety/gastronomy of the food basket, preferable hypermarkets, cost of food products, take-away food, dining outside home and other questions aiming to develop an acceptable food basket as one of the major factors affecting the level of adherence to a specific diet(Reference Nelson, Dick and Holmes24). Moreover, a typical weekly diabetic food menu was attached with the questionnaire to ensure a better understanding of the proposed DHFB by participants.
Cost and affordability
The pricing of DHFB was based on Chrysostomou and Andreou(Reference Chrysostomou and Andreou18). Most of the data were retrieved from the website of the Ministry of Energy, Commerce, Industry and Tourism of the Republic of Cyprus(25). Moreover, using the Consumer Price Index (CPI) provided by the Statistical Service of Cyprus, all products were adjusted to 2019 prices.
Determining the adequacy of income levels requires their evaluation against a benchmark(Reference Friel, Walsh and McCarthy13). An appropriate benchmark for assessing food affordability in this study is the minimum income threshold as defined by the Guaranteed Minimum Income (GMI) scheme. The scheme was introduced in Cyprus in 2014 following an important social policy reform aiming at a more efficient and targeted social protection, thereby providing relief to households vulnerable to income deprivation and social exclusion(Reference Koutsampelas26). GMI is a top-up benefit, defined as the difference between a minimum income threshold and family income. Simply stated, if a family’s income falls below the minimum income threshold, the state steps up to fill the difference, provided a series of other eligibility conditions are satisfied. The minimum income threshold represents the minimum income necessary to ensure recipients’ access to a basket of goods and services corresponding to the minimum socially accepted standard of living. The value of this basket, calculated by the Ministry of Labour, Welfare and Social Insurance using a reference budget methodology, is currently set at €480 per month for a single adult and increases for larger recipient units (by 50 % for an additional adult and by 30 % for an additional child)(Reference Koutsampelas26).
The affordability of DHFB was measured by calculating the cost of the food basket as a percentage of household income and occurs when households spend ≤30 % of their income on food costs(Reference Ward, Verity and Carter16). Thus, the cost of each basket as a percentage of GMI is used as a yardstick of affordability.
Results
Participants’ socioeconomic background
As already mentioned, the sample consisted of 422 adults (59 % men and 41 % women). The average age was 65·3 years (66·5 for men and 63·5 for women) due to the fact that the sample was derived from public hospital visitors, typically of older age. Although the DHFB was developed for adults around 40 years, this would not introduce any kind of bias since the basic nutritional requirements (at a macronutrient level) might not differ among younger and older adults (up to 65 years), thereby not significantly affecting the content and cost of DHFB. About 35 % of participants had only completed primary education or less, 49·5 % had completed up to upper secondary education (including post-secondary non-tertiary level) and only 15·2 % were highly educated (i.e. holding a bachelor’s degree or above). The average educational attainment was lower among women (e.g. 44·8 % of female participants had completed primary education or less), reflecting the lower participation of women in education among the older cohorts of the population. As it can be expected, the majority of participants were pensioners (66·4 %); yet, there was also a significant share of employed (17·1 %) and unemployed persons (6·4 %). Finally, other economic activities included mostly housewives (14·9 % among female respondents). It is worth mentioning that a significant proportion of participants (28·0 %) reported very low annual family income (<€11 000 per annum), with another 38·4 % reporting rather low family income (between €11 001 and €20 000), and only 28·0 and 3·8 % reporting family income between €20 001 and €40 000 and >€40 001, respectively. The above are described in detail in Table 1.
Table 1 Participants’ socioeconomic backgrounds by gender
* Secondary education also includes the non-tertiary post-secondary level.
† ‘Other’ category in economic activity includes housewives.
Diabetic Healthy Food Basket v. Healthy Food Basket
A DHFB is described in detail in Table 2. The majority of participants (89·4 %) considered DHFB as acceptable, while about 95 % mentioned that DHFB has sufficient taste and consists of a variety of food items. DHFB consists of eight food categories, similar to the Cypriot HFB. The distribution of food groups remained the same for DHFB but differed in food quantity due to the effect of gender. Although food categories remained the same, specific food items were removed and others were replaced following ADA guidelines and information collected through the questionnaires(Reference Evert, Boucher and Cypress1). A more detailed description of adaptations in DHFB compared with HFB follows.
Table 2 Components of the Diabetic Healthy Food Basket for each member of the household based on the Cypriot National Guidelines for Nutrition and Exercise(22), American Diabetes Association(Reference Evert, Boucher and Cypress1) guidelines for diabetes, and information retrieved from questionnaires
Compared with the Cypriot HFB, some food items were removed. Concerning the Liquid group, all alcohol drinks were removed (wine and beers). Based on the ADA guidelines, alcohol consumption may place people with diabetes at an increased risk of delayed hypoglycaemia, especially if taking insulin or insulin secretagogues(Reference Evert, Boucher and Cypress1). Regarding the Grains group, all food items remained the same but were replaced by whole wheat so as to ensure adequate consumption of fibre and whole grains following the ADA guidelines(Reference Evert, Boucher and Cypress1). In regard to this, several studies using low-glycaemic-index eating patterns have demonstrated improved glycaemic control(Reference Wheeler, Dunbar and Jaacks27,Reference Jenkins, Srichaikul and Kendall28) . Moreover, in the same group, weekly number of portions for specific food items has been reduced. This change was based on the ADA guidelines, which support that an MNT provided by a RD should reduce daily energy intake (232–710 kcal/d) to provide modest weight loss and prevention of weight gain followed by a potential effect on glycaemic profile(Reference Evert, Boucher and Cypress1). In particular, bread, wholegrain (day: breakfast, snack, side dish) was reduced from seven portions per week to three portions per week. Also, breakfast cereals, not sweetened were reduced from seven portions per week to four portions per week. In the same group, portions for pasta were reduced from three to one portion per week. In the Vegetables group, frozen unprepared vegetables and vegetable juice were replaced with fresh vegetables for being good sources of vitamins and minerals. This change was also based on the ADA guidelines that promote the consumption of fresh fruits and vegetables as these food items have shown significant improvements in glycaemic control(Reference Evert, Boucher and Cypress1,Reference Wheeler, Dunbar and Jaacks27) . In the Fruits group, all food items remained the same and only weekly portion of fruit juice was changed from four to three portions per week. In the Dairy group, only cheese was changed from mature to low-fat cheese and the weekly portions were increased from two to three since individuals with diabetes should moderate their fat intakes to be consistent with their goals to lose or maintain weight(Reference Evert, Boucher and Cypress1). In the Meat/Fish/Eggs group, only fish canned was replaced with fresh fish, whereas the number of portions remained the same. This change was based on the recommendation for the general population to consume fish (particularly fatty fish) at least two times (two servings) per week, which is also appropriate for people with diabetes(Reference Evert, Boucher and Cypress1). In the Fat group, weekly portions of nuts were reduced from seven to four. Finally, in the Residuals group, weekly portions of Choco were increased from one to two and replaced with dark choco (participants’ requirement). Ice cream was removed (RD’s recommendation). Additionally, participants asked for more sweets in DHFB, to which the RD disagreed. Moreover, jam and honey were removed from the basket (participants’ requirement). Salt and cold sauces (ketchup) were also removed following the RD’s recommendation. In regard to salt consumption, a Cochrane review of randomized controlled trials has found that decreasing sodium intake reduced blood pressure and improved cardiovascular risk in those with diabetes(Reference Suckling, He and Macgregor29). Therefore, salt was completely removed from the basket. Notably, a great majority of participants asked to add more Cypriot traditional foods within the food basket, but the RD disagreed since these foods consist of high amounts of saturated fats that have negative health effects(30). However, all participants agreed with the final version of DHFB, and thus, after all adjustments, the DHFB could be considered acceptable for its population.
Cost and affordability of the Diabetic Healthy Food Basket
Table 3 shows the monthly budget required for each food item included in DHFB. The total budget for DHFB is lower for both households (i.e. single woman and single man) compared with HFB. The total monthly budget for a diabetic woman is about 15 % lower (25·68 Euros less) compared with HFB, and the relative percentage for a diabetic man is about 16 % (37·58 Euros less) lower. Also, the total required monthly budget for a diabetic woman is around 30 % lower (60·32 Euros less) compared with that of a diabetic man.
Table 3 Cost of the Diabetic Healthy Food Basket v. Healthy Food Basket, monthly amounts in Euros, July 2019
Table 4 presents the food values corresponding to DHFB and HFB for women and men in Cyprus in terms of GMI. Line 1 presents the corresponding GMI values. Lines 2 and 4 present the monthly costs for DHFB and HFB for women and men, respectively. The monthly cost for women is lower compared with that for men in both lines. Finally, lines 3 and 5 show the proportions of GMI that needs to be spent on DHFB and HFB, respectively. For low-income adults in Cyprus receiving GMI, the proportion of income that would be spent on DHFB ranges from around 30 % to 42 % for women and men, respectively. In addition, the relative proportions for HFB (line 5) are higher compared with that for DHFB for both genders. However, the difference in affordability rates between DHFB and HFB for women is lower than that of men (around 5 v. 8 %, respectively).
Table 4 Guaranteed Minimum Income (GMI), Diabetic Healthy Food Basket (DHFB) and Healthy Food Basket (HFB) for Cypriot adults
Discussion
Nutrition has been the cornerstone of therapy to enable persons with diabetes to manage their chronic disease, prevent complications and provide a good quality of life(Reference Franz, Splett and Monk9). Although the need for data on the outcomes of costs of diabetes treatment, including nutrition therapy, has been expressed repeatedly(Reference Franz, Splett and Monk9), relative studies are still scarce. Therefore, this study aimed to develop an acceptable Cypriot DHFB and provide information relating to the cost and affordability of this basket among the low-income population. The current study shows that the cost of Cypriot DHFB is lower compared with Cypriot HFB, meaning that nutritional treatment based on the practice guidelines for diabetes could result in cost savings for these patients.
In regard to the affordability of Cypriot DHFB, results seemed to be more promising compared with Cypriot HFB and other disease-based food baskets. Results of the current study show that DHFB is more affordable compared with HFB among low-income Cypriots. Notably, for low-income women, it seems that the purchase of DHFB could be defined as affordable since <30 % of their income is required for purchasing healthy food(Reference Ward, Verity and Carter16). However, the rate of affordability is marginal (29·7 %), indicating that the risk of experiencing food insecurity still exists. Hence, compared with HFB, it seems that purchasing DHFB is more affordable (29·7 v. 35 %) for a low-income diabetic Cypriot woman (Table 4). On the other hand, the purchase of DHFB for an adult (low-income) man with diabetes is not affordable, despite affordability is better compared with HFB (42·3 v. 50·1 %) (Table 4). Thus, it could be assumed that both household types are at risk of experiencing food insecurity due to low affordability, but the risk is higher for diabetic men. Moreover, the risk of experiencing food insecurity is lower for DHFB compared with HFB for both types of households. Based on the above findings, it could be supported that nutrition therapy that follows practice guidelines for diabetes is more affordable compared with the basic nutrition therapy.
Findings of the current study are in contrast with previous findings relating to the food baskets developed for patients with other chronic diseases requiring MNT. Particularly, a recent study in Cyprus has shown that for low-income people diagnosed with celiac disease and receiving GMI, the proportion of income that must be spent on GFHFB ranges from around 42 to 60 %. Particularly, GFHFB was 33·6 and 47 Euros/month more expensive compared with HFB for women and men, respectively(Reference Chrysostomou, Andreou and Andreou19). Thus, comparing the three food baskets (DHFB, GFHFB, HFB) developed for the Cypriot population, it seems that DHFB is the most cost-efficient nutrition therapy for both household types (142·61/202·93 euros, 201·2/285·8 euros, 168·29/240·51 euros for women and men, respectively). Moreover, comparing disease-based food baskets (DHFB and GFHFB), it seems that DHFB is a more cost-efficient therapy compared with GFHFB, and the cost may be 29 % lower for both household types (woman and man). In addition, affordability rates for low-income people range accordingly. Better affordability is shown for DHFB, following HFB, and then higher affordability is shown for GDHFB (41·92, 59·54 % for woman and man, respectively)(Reference Chrysostomou and Andreou18,Reference Chrysostomou, Andreou and Andreou19) .
MNT provided by an RD is a key complement to traditional medical interventions in several chronic diseases(Reference Morris and Wylie-Rosett8). Regardless of medical and clinical benefits, studies have shown that nutrition interventions that follow practice guidelines, such as the MNT, provide a reasonable economic investment(Reference Franz, Splett and Monk9) Thereafter, ADA guidelines support that diabetes nutrition therapy can result in cost savings(Reference Evert, Boucher and Cypress1). The current study confirms the above statement regarding the effectiveness of diabetes nutrition therapy(Reference Evert, Boucher and Cypress1). Particularly, this study shows that developing a national DHFB in line with evidence-based nutrition recommendations and considering the population’s needs and preferences in the treatment of diabetes is more cost-efficient compared with basic nutrition therapies, but still the issue of affordability remains among the low-income population, mainly among diabetic men. Although affordability has improved compared with other nutrition therapies, it still requires attention.
The implications for public policy are straightforward. Although DHFB is less costly than HFB, it still remains unaffordable or marginally affordable for the GMI-supported population (or for any person facing income deprivation). Although not examined explicitly in this study, the problem of affordability is likely to deteriorate for larger family units, as food consumption is characterised by limited household economies of scale (i.e. adding one adult to the household would almost double the cost of food basket, while it might increase welfare payments only by 50 %). Thus, the specialised nutritional needs of the diabetic population emerge as an important policy concern not only for reforming income schemes but also for the formulation of public health policy. This is crucial for an additional reason. As shown in the literature, low income is associated with a higher prevalence of diabetes and diabetes-related complications(Reference Rabi, Edwards and Southern31). This was evidenced in our data where 28 % of participants reported living off very low annual family income. Further, 26 % of participants reported that the financial situation of their family is very strongly (10 %) and strongly (16 %) affected by their health condition, highlighting the hardship in maintaining a healthy nutrition in the presence of income vulnerability.
Overall, the findings of this study should encourage similar studies in other countries in expectation of more useful information relating to the cost of DHFB in different socioeconomic contexts. Notably, MNT may not always be cost-efficient since this depends on the disease. However, in regard to diabetes, each country should assess the cost of a national DHFB and examine the possibility of financially supporting this diet, especially for the low-income population. The expected benefits of this initiative on improving public health, reducing health inequalities and promoting economic efficiency would be unambiguously large.
Acknowledgements
Acknowledgements: The authors would like to acknowledge the medical staff of the public hospital of Nicosia and Limassol for their valuable collaboration and contribution to this study. Financial support: None. Conflict of interest: None. Authorship: All authors contributed to the conduct and reporting of the work. All authors commented on drafts and read and approved the final manuscript. The corresponding author attests that all listed authors meet the authorship criteria and that none meeting the criteria have been omitted. Ethics of human subject participation: This study was conducted according to the guidelines laid down in the Declaration of Helsinki, and all procedures involving study participants were approved by the Cypriot Bioethics Committee and the Office of the Commissioner for Personal Data Protection. Written informed consent was obtained from all subjects/patients.
