Podcast available onlineat www.jneb.org Research Brief

The Cost of a Healthier Diet for Young Children With Type1 Diabetes MellitusSusana R. Patton, PhD, CDE1; Kathy Goggin, PhD2; Mark A. Clements, MD, PhD3

ABSTRACT

Objective: This study used a market-basket approach to examine the availability and cost of a standardfood shopping list (R-TFP) vs a healthier food shopping list (H-TFP) in the grocery stores used by a sampleof 23 families of young children with type 1 diabetes mellitus (T1DM).Methods: The researchers used frequency counts to measure availability. The average cost of the R-TFPand H-TFP was compared using paired t test.Results: Small or independent markets had the highest percentage of missing foods (14%), followed bychain supermarkets (3%) and big box stores (2%). There was a significant difference in average cost for theR-TFP vs the H-TFP ($324.71 and $380.07, respectively; P < .001).Conclusions and Implications: Families may encounter problems finding healthier foods and/or incurgreater costs for healthier foods. Nutrition education programs for T1DM need to teach problem solvingto help families overcome these barriers.Key Words: nutrition, behavior, child, diabetes, food costs (J Nutr Educ Behav. 2015;47:361-366.)

Accepted March 15, 2015.

INTRODUCTION

It is standard in nutrition educationfor type 1 diabetes mellitus (T1DM)to instruct patients and familiesto reduce fat and increase fruits, vege-tables, and whole grains in theirdiets.1,2 However, despite these rec-ommendations, from the availableresearch, it is known that many youthswith T1DM do not consume ahealthful diet.3-5 A common problemreported by parents is that providingmore nutritious foods to their child isdifficult, time-consuming, and expen-sive.6 Cost and availability are estab-lished barriers to healthful eating inthe general population and havebeen quantified by market-basketstudies.7-10 Similarly, in youths withT1DM, 1 study has shown a relationbetween higher diet costs and a

healthier diet.11 However, that studyrecruited youths aged 8–18 yearsand diet costs were estimated basedon price information from 2 onlinenational supermarkets comparedwith actual stores where familiesroutinely shopped. It remains un-known whether there is a higher costto healthier eating for families ofyoung children with T1DM, a subsetof patients who likely eat a majorityof their meals at home and typical-ly have parent supervision of theirmeals, which creates an oppor-tunity for healthier eating.12,13 Also,because the prior study of youthswith T1DM was not a market-basketstudy, it is not known whether fam-ilies of youths with T1DM face chal-lenges in finding healthier foodoptions in their local stores. Thus,this market-basket study sought to

examine the physical availability andthe cost of healthier foods in thestores that parents of young childrenwith T1DM reported using for routinefood shopping in Northeastern Kan-sas and Western Missouri. Market-basket studies are commonly used toexamine the physical availability andcost of foods that make up a nutriti-ous diet.7,14,15 These studies use asurvey approach to gather databased on a standard shopping list instores identified based on the studysample (ie, young children or urbanneighborhood). Following publishedmethodology,7 the researchers usedthe US Department of Agriculture’sThrifty Food Plan (R-TFP)14 and amodified healthier version of theThrifty Food Plan (H-TFP)7 to deter-mine food prices for 2 standard shop-ping lists. The researchers specificallysought to answer the following ques-tions: (1) What is the physical avail-ability of healthier food options inthe stores that parents of young chil-dren with T1DM use for routine foodshopping? (2) Is there a difference inthe price of the R-TFP vs the H-TFPat these stores? Thus, this study pro-vides valuable data to establish thecost and physical availability ofhealthful food options for a sampleof families of children with T1DMthat will ultimately lead to better indi-vidualization of diabetes nutrition

1Department of Pediatrics, University of Kansas Medical Center, Kansas City, KS2Health Outcomes Research, Children’s Mercy Hospital and Clinics, Kansas City, MO3Division of Endocrinology, Children’s Mercy Hospital and Clinics, Kansas City, MOConflict of Interest Disclosure: The authors’ conflict of interest disclosures can be found onlinewith this article on www.jneb.org.Address for correspondence: Susana R. Patton, PhD, CDE, University of Kansas MedicalCenter, 3901 Rainbow Blvd, MS 4004, Kansas City, KS 66160; Phone: (913) 588-6323;Fax: (913) 588-2253; E-mail: spatton2@kumc.edu�2015 Society for Nutrition Education and Behavior. Published by Elsevier, Inc. All rightsreserved.http://dx.doi.org/10.1016/j.jneb.2015.03.006

Journal of Nutrition Education and Behavior � Volume 47, Number 4, 2015 361

UNDER EMBARGO UNTIL JULY 8, 2015, 12:01 AM ET

education and new curricula toproblem-solve challenges created bythese barriers.

METHODSProcedure

The researchers recruited families froma pediatric diabetes clinic in the Mid-western US to participate. Familieswere eligible if they had a child withT1DMwho was aged 1–6 years, at least6 months beyond his or her T1DMdiagnosis, and on an intensive insulinregimen. A total of 27 families initiallyagreed to participate but 3 familieswere lost because of illness and 1 couldnot be reached to schedule a study visit(85% participation rate). All study pro-cedures were approved by the Chil-dren's Mercy Hospital InstitutionalReview Board before subject recruit-ment; parents provided written in-formed consent before participating.

Parents provided demographic in-formation and the name and locationof the primary store where they com-plete their weekly food shopping dur-ing a home study visit. Mastersstudents in dietetics whowere blinded

to the study questions were then dis-patched to each of the stores identi-fied by parents to collect the pricesfor 164 food items on a standard listthat included all items on the R-TFPand H-TFP. Table 1 provides a samplelist.7,14 Students were instructed torecord the lowest non-sale price perunit for each food item. Once thesedata were collected, a research assis-tant double-checked the lists formissing or potentially inaccurate pri-ces and corrected them as needed.Potentially inaccurate prices were sus-pected if the price listed for an itemwas > 3 SD from the mean price ofthat item from similar stores. If aninaccurate price was suspected or amissing price was found, the protocolwas for the research assistant to callthe store to check the price. In caseswhere a missing price was due to amissing food (4%; 151 of 3,772), asubstitution was made based on themean price of that food from similarstores. The validated food prices werethen entered into a spreadsheet andthe store’s total costs for the R-TFP vsthe H-TFP were automatically calcu-lated for each family based on a presetformula. As an additional descriptive

variable, children’s most recent gly-cated hemoglobin A1c (HbA1c) valuewas collected from their medical re-cord. These values were obtained usingthe DCA 2000þ Analyzer (Bayer Cor-poration, Tarrytown, NY; normativereference ranges, 4.5% to 5.7%). Themean number of days from the timeof children’s first study visit and theirreportedHbA1cvaluewas17�14days.

Measures

To assess family socioeconomic status,the researchers used the Hollingshead4-Factor Index, which computes aclass score based on parents’ maritalstatus, education level, job title, andemployment status (full or part-time).16

Thrifty Food Plan

The researchers selected the R-TFPbecause it included food items meantto describe a typical 2-week grocerylist for a family of 4 living on amodestfood budget or participating in theSupplemental Nutrition AssistanceProgram. Also, although the R-TFPprovides a diverse grocery list thatcan enable a family to meet USDepartment of Agriculture dietaryintake guidelines, there is room foradditional healthier food substitu-tions, which can increase its healthvalue. For example, Jetter and Cas-sady7 identified healthier substitutesfor the original dairy, meats, cannedfruit, fats, breads, and other grainproducts, leading to an H-TFP shop-ping list that had about 4 times theamount of fiber and one fifth thegrams of total fat as the R-TFP. Fooditems that did not change from theR-TFP to the H-TFP included freshfruits and vegetables, eggs, beans,and spices and herbs. Although datafor the current study were collectedfrom families of young children withT1DM, virtually all families reportedonly minimal use of sugar-free prod-ucts, based on a study questionnaire.Thus, no additional substitutions forsugar-free products were made. Also,juice cocktail and 100% juice were re-tained in the R-TFP and the H-TFP,respectively, because when they wereasked, 90% of parents indicated thatthey may serve juice when treatinghypoglycemic episodes. Similar stores

Table 1. Example Items in the Regular Thrifty Food Plan and Healthier Thrifty FoodPlan

Food Item

Baskets

Regular ThriftyFood Plan

Healthier ThriftyFood Plan

Breads/grains Refined Whole grain

Canned fruit Heavy or light syrup Light syrup or juice

Canned vegetables Standard Low sodium

Chicken Thighs Breast

Cooking oil Vegetable Canola

Flour White Whole wheat

Frozen fish Filets with breading Filets, no breading

Frozen potatoes Frozen french fries Frozen potatoes

Ground meat # 20% fat # 10% fat

Milk 1%, whole 1%, nonfat

Pasta Enriched Whole wheat

Rice White Brown

Salad dressing Regular Low fat

Spreads Jelly Real fruit spread

Tomato soup Regular Low sodium

Note: Items included in the Healthier Thrifty Food Plan were previously modified byJetter and Cassady.7

362 Patton et al Journal of Nutrition Education and Behavior � Volume 47, Number 4, 2015

were grouped based on type: chain su-permarkets (stores > 20,000 squarefeet), small and independent markets(< 20,000 square feet), and big boxstores (stores > 20,000 square feetthat sold bulk items but did notcharge a membership fee).7

The total price for each food wascalculated by multiplying the speci-fied quantity of each food based onthe R-TFP or H-TFP by the lowestunit price as collected from each store.Then, to calculate the final marketprice for the R-TFP and the H-TFP,the total costs for items on each listwere summed. Also, the final marketprices for each of the 4 componentsof the R-TFP and H-TFP (ie, fruit andvegetables, meat, grains, dairy) werecalculated by summing the items spe-cific to each of these components. Theresearchers used paired t tests tocompare final market costs for the R-TFP and H-TFP as well as each of the4 component scores (adjusted a ¼.01). Finally, descriptive statisticswere used to describe the sample, thetypes of stores used by families, andthe physical availability of selectedhealthier foods by store type.

RESULTS

The sample consisted of 23 familieswith young children with T1DM.Mean age of young children was 4.6� 1.3 years; 43.5% were female and78.3% of parents identified their childas white. Young children had a meanHbA1c level of 7.8% � 1.0% andwere 2.0 � 1.5 years post-diagnosis;87% of children used an insulinpump. Parents’ mean age was 35.7 �5.1 years. A total of 91% of partici-pating parents were mothers, 82.6%of parents were married, 78.3% of par-ents reported at least some college,and 82.6% of families reported mid-dle- to upper middle–class socioeco-nomic status.

Small and independent marketshad a higher percentage of missingfoods (14.3% � 11.9%) comparedwith chain supermarkets (3.4% �2.8% of foods missing) and big boxstores (2.5% � 1.1% of foods missing)(Table 2). Fourteen families reportedshopping at a chain supermarket, 6 re-ported shopping at a big box store,and 3 reported shopping at a smallor independent market. There were

20 unique stores surveyed, distributedwidely across the metropolitan areaand within a mean 29-mile radius ofthe main clinic (Figure).

Consistent with parents’ percep-tions, the H-TFP was more expensivethan the R-TFP (t[22] ¼ �17.01; P <.001). The average cost for the R-TFPwas $324.71 vs $380.07 for the H-TFP, which translated into a meandifference of $57.62 � $14.02 (range,$31.10 to $87.87) per basket. Health-ier alternatives tended to cost morefor all of the components, but thegreatest differences in costs were

observed for protein and grains. Forprotein, the average increase in costfor the H-TFP was $23.06 � $9.29whereas for grains the average in-crease in cost was $15.48 � $5.91.Simple correlations revealed no asso-ciation between families’ Hollings-head Socioeconomic Score and theprice they might pay for the R-TFP (r¼ �.06; P ¼ .79) or the H-TFP (r ¼�.07; P ¼ .73) at their respectivestore. Categorical comparisons werenot possible because a low frequencyof families fell below the middle-classsocioeconomic status.

Table 2. Stores Missing Specific Healthier Foods, Grouped by Store Type, n

Missing Foods

Small/Independent

MarketsChain

SupermarketsBig BoxStores Total

Catsup, low sodium 1 1 0 2

Cheddar cheese, low fat 2 2 0 4

Chicken bouillon, low sodium 1 0 0 1

Chicken fryer, no skin 2 5 0 7

Chuck roast, lowest fat 2 10 0 12

Cod filets 2 7 4 13

Chocolate drink mix, sugar free 1 1 0 2

Egg noodles, yolkless 1 0 0 1

English muffins, whole wheat 2 0 0 2

Flour, whole wheat 1 0 0 1

Garbanzo beans, canned 1 0 0 1

Great northern beans, canned 1 0 0 1

Green beans, canned low sodium 1 0 0 1

Hamburger bun, whole wheat 1 0 0 1

Lima beans, dry 2 0 0 2

Macaroni, whole wheat 1 2 0 3

Pancake syrup, light 1 0 0 1

Pears, canned light syrup 1 0 0 1

Popsicles, sugar free 1 0 0 1

Pork, ground low fat 1 1 2 4

Real fruit spread 2 0 0 2

Rolls, whole wheat 1 6 0 7

Spaghetti sauce, low sodium,low sugar

1 2 0 3

Soy sauce, low sodium 1 0 0 1

Tomato paste, low sodium 1 0 0 1

Tomato soup, low sodium 1 3 2 6

Turkey breast, ground 2 1 0 3

Turkey ham (deli) 1 7 3 11

Note: Data reflect missing healthier foods from 3 small or independent markets, 14chain supermarkets, and 6 big box stores.

Journal of Nutrition Education and Behavior � Volume 47, Number 4, 2015 Patton et al 363

DISCUSSION

This study explored the physical avail-ability and cost of healthier foods us-ing a market-basket design in asample of families with young chil-dren with T1DM. With respect to thephysical availability of healthierfoods, the results showed that smalland independent markets tended tohave more missing foods than chainsupermarkets and big box stores. Inthe literature, problems with thephysical availability of healthier foodshave been identified for small and in-dependent markets located in lower-income urban neighborhoods and insome rural communities, leading to afood desert for some families thatstruggle with gaining access to ahealthful diet.7,17-19 In this study, 3families reported using a small orindependent market for weekly foodshopping. Two of these familiesreported shopping in upscaleindependent stores that featuredprimarily local products and werelocated in middle- to higher-incomeneighborhoods. However, 1 of thesefamilies lived in a small rural townwith 1 independent market, and thenearest larger store, a big box store,

was located about 25 miles away.Thus, similar to the results of pastfood desert studies,7,17-19 it is likelythat this family faced a significantbarrier to more healthful eating as aresult of limited access. It isestimated that US families make anaverage of 2.2 trips to a grocery storeper week.20 Therefore, for this familyto shop at the nearest big box store,it would add both significant timeand expense to their food shopping,which suggests that the potentialnegative effects of a food desert mayalso exist outside lower-income urbancommunities.

Specific to cost, the mean differ-ence between the R-TFP and H-TFPwas $57.62/basket. Thus, over 1 year,the difference in cost for a familyshopping from the H-TFP vs the R-TFP would be $1,498.12. In 2013,Americans spent a mean of $3,997on foods purchased at a grocery orother store and prepared by the con-sumer (food-at-home expenses).21

Based on these values, the highercost of the H-TFP would be about37.5% of a family’s food-at-home ex-penses. This is consistent with find-ings from an earlier study, whichreported the higher cost of the health-

ier basket to be between 35% and 40%of a family’s food-at-home expenses.7

This finding is also generally consis-tent with a recent study conductedin older youths with T1DM, whichfound that higher food costs wereassociated with a healthier diet.11 Par-ents commonly report the higher costof fresh fruits and vegetables.8,22-24

Although the cross-sectional natureof this study’s data does not takeinto account the volatility of vege-table and fruit pricing in the globalmarkets, based on the R-TFP, familieswould spend approximately $50.22on fresh fruit and vegetables every 2weeks. This is equivalent to 16% ofthe total average price of the R-TFPand 13% of the total average price ofthe H-TFP. However, healthier ver-sions of canned fruits (canned withjuice) and vegetables (lower sodium)or frozen products without added sau-ces, sodium, or sugar may offer a moreaffordable option. Overall, thesehealthier canned or frozen fruits andvegetables were widely available, andin some cases store brand optionswere available, leading to additionalcost savings for consumers. Based onthe surveys, the mean 2-week cost ofhealthier canned fruits and vegetableswas $13.55, which is equivalent to0.3% of both the total average pricesof the R-TFP and the H-TFP. Themean 2-week cost of healthier frozenfruits and vegetables was $31.72, orthe equivalent of 10% of total averageprices for the R-TFP and 8% of the to-tal average price of the H-TFP. Thus, togive parents an affordable alternativeto fresh produce, nutrition educationprograms should consider teachingparents about healthier canned andfrozen fruits and vegetables, and re-view how to measure and count car-bohydrates in these foods, becausecarbohydrate counts will be higherin canned fruits.

This study presents new informa-tion regarding differences in the costof a healthier diet in a sample of fam-ilies with young children with T1DM.Limitations of this study include itssmall and homogeneous sample.Although the ethnic, racial, and socio-economic composition of the sampleis similar to the clinic from which itwas drawn, the results may not gener-alize to children from other ethnic,racial, or socioeconomic backgrounds.Another limitation is the location of

Figure. Distribution of store surveyed relative to the main diabetes clinic wherefamilies were recruited in the Kansas City, KS, metropolitan area. Blue star indi-cates main diabetes clinic; red squares, stores. A total of 18 stores are shown;3 stores were duplicates (ie, 2 different families happened to shop at the samestore).

364 Patton et al Journal of Nutrition Education and Behavior � Volume 47, Number 4, 2015

the study, which recruited familiesfrom a diabetes clinic located in amoderately sized city in the Midwest-ern US. Because cost of living canalso affect local food prices, futuremarket-basket surveys are needed todetermine food prices and physicalavailability in other cities or regionsin the US. Finally, the study method-ology may be a limitation. This studyused 2 standard grocery lists in amarket-basket design to documentfood physical availability and cost.This methodology allowed for com-parisons across individual familiesand store types and to ensure that awide variety of foods were surveyed.An alternative strategy would havebeen to generate individual shoppinglists for families based on diet diary in-formation. This approach would haveallowed for a better estimate of thetrue difference in costs for familiesbased on their individual buyinghabits. Thus, the results provideinsight into the degree of cost differ-ences families with young childrenwith T1DM could face when eatinghealthier foods, but not the specificcost difference they encounter in theirweekly shopping.

IMPLICATIONS FORRESEARCH ANDPRACTICE

The results show that a healthier mar-ket basket can cost 18%more than thestandard basket. Moreover, familiescan face barriers in finding specifichealthier foods at their local stores.In addition to educating parents andyouths regarding recommendationsfor healthful eating, diabetes nutri-tion programs need to address barriersto healthful eating that are created bythe marketplace and food environ-ment in which their patients areliving. Some individualized nutritioncounseling strategies that might helpinclude reframing food purchases interms of nutrition per dollar to teachfamilies how to shop for healthfulfoods at the lowest possible price,providing recipes and teaching fam-ilies how to cook lower-cost substi-tutes for higher-priced foods, andproviding information on local storesthat offer a wide selection of healthfulfoods. In short, collaborative problemsolving regarding marketplace barriers

to healthful eating has the potentialto increase the likelihood that parentswill purchase more healthful foodseach week and decrease parents’ feel-ings of frustration or stigma relatedto their food shopping practices.

ACKNOWLEDGMENTS

This research was supported in part bygrant R03-DK090288 (to SRP) fromthe National Institutes of Health/Na-tional Institute of Diabetes and Diges-tive and Kidney Diseases. The authorsthank Ms Allison Wilcox, Ms AshleyMoore, and Ms Courtney Moore ofthe University of Kansas Medical Cen-ter for assistance in conducting themarket-basket surveys.

REFERENCES

1. Bantle JP, Wylie-Rosett J,Albright AL, et al. Nutrition recom-mendations and interventions for dia-betes: a position statement of theAmerican Diabetes Association. Dia-betes Care. 2008;31(suppl 1):S61-S78.

2. Silverstein J, Klingensmith G,Copeland K, et al. Care of childrenand adolescents with type 1 diabetes: astatement of the American Diabetes As-sociation. Diabetes Care. 2005;28:186-212.

3. Helgeson VS, Viccaro L, Becker D,Escobar O, Siminerio L. Diet of adoles-cents with and without diabetes:trading candy for potato chips?DiabetesCare. 2006;29:982-987.

4. Mayer-Davis EJ, Nichols M, Liese AD,et al. Dietary intake among youth withdiabetes: the SEARCH for Diabetes inYouth Study. J Am Diet Assoc. 2006;106:689-697.

5. Patton SR, Dolan LM, Chen M,Powers SW. Dietary adherence andmealtime behaviors in young childrenwith type 1 diabetes on intensive insulintherapy. J Acad Nutr Diet. 2013;113:258-262.

6. Mehta SN, Haynie DL, Higgins LA,et al. Emphasis on carbohydrates maynegatively influence dietary patterns inyouth with type 1 diabetes. DiabetesCare. 2009;32:2174-2176.

7. Jetter KM, Cassady DL. The availabil-ity and cost of healthier food alterna-tives. Am J Prev Med. 2006;30:38-44.

8. Drewnowski A. The cost of US foodsas related to their nutritive value. Am JClin Nutr. 2010;92:1181-1188.

9. Green R, Cornelsen L, Dangour AD,et al. The effect of rising food priceson food consumption: systematic re-view with meta-regression. BMJ.2013;346:f3703.

10. Rao M, Afshin A, Singh G,Mozaffarian D. Do healthier foods anddiet patterns cost more than less healthyoptions? A systematic review and meta-analysis. BMJ Open. 2013;3:e004277.

11. Nansel T, Haynie D, Lipsky L, Mehta S,Laffel L. Little variation in diet cost acrosswide ranges of overall dietary qualityamong youth with type 1 diabetes. JAcad Nutr Diet. 2015;115:433-439.

12. Ziegler P, Briefel R, Ponza M,Novak T, Hendricks K. Nutrient in-takes and food patterns of toddlers’lunches and snacks: influence of loca-tion. J Am Diet Assoc. 2006;106(suppl1):S124-S134.

13. Tatlow-GoldenM,HennessyE,DeanM,Hollywood L. ‘‘Big, strong and healthy’’:young children’s identification of foodand drink that contribute to healthygrowth. Appetite. 2013;71:163-170.

14. US Department of Agriculture Centerfor Nutrition Policy and Promotion.Thrifty Food Plan. http://www.cnpp.usda.gov/sites/default/files/usda_food_plans_cost_of_food/TFP2006Report.pdf.Accessed April 7, 2015.

15. US Department of Health and HumanServices, Department of Agriculture.Dietary guidelines for Americans 2005.Washington, DC: Government Print-ing Office; 2005.

16. Hollingshead AB. Four Factor Index ofSocial Status. New Haven, CT: YaleUniversity Press; 1975.

17. DiSantis KI, Grier SA, Odoms-Young A, et al. What ‘‘price’’ meanswhen buying food: insights from amultisite qualitative study with BlackAmericans. Am J Public Health. 2013;103:516-522.

18. Steenhuis IH, Waterlander WE, deMul A. Consumer food choices: therole of price and pricing strategies. Pub-lic Health Nutr. 2011;14:2220-2226.

19. Zenk SN, Powell LM, Rimkus L, et al.Relative and absolute availability ofhealthier food and beverage alternativesacross communities in the United States.Am J Public Health. 2014;104:2170-2178.

20. Statista Statistics Portal. Food & Nutri-tion: Grocery Shopping: U.S. con-sumers’ weekly trips per household2006-2012. http://www.statista.com/statistics/251728/weekly-number-of-us-grocery-shopping-trips-per-household/.Accessed August 22, 2014.

Journal of Nutrition Education and Behavior � Volume 47, Number 4, 2015 Patton et al 365

21. US Department of Labor, Bureau ofLabor Statistics. Consumer Expendi-ture Survey. http://www.bls.gov/cex/2013/csxann13.pdf. AccessedApril 7, 2015.

22. Stark LJ, Mulvihill MM, Powers SW,et al. Behavioral intervention toimprove calorie intake of children

with cystic fibrosis: treatment versuswait list control. J Pediatr GastroenterolNutr. 1996;22:240-253.

23. Stark LJ, Powers SW, Jelalian E,Rape RN, Miller DL. Modifyingproblematic mealtime interactions ofchildren with cystic fibrosis and theirparents via behavioral parent

training. J Pediatr Psychol. 1994;19:751-768.

24. Patton SR, Odar C, Midyett LK,Clements MA. Pilot study results for anovel behavior plus nutrition interven-tion for caregivers of young childrenwith type 1 diabetes. J Nutr Educ Behav.2014;46:429-433.

366 Patton et al Journal of Nutrition Education and Behavior � Volume 47, Number 4, 2015