Preparation of Gluten-Free Foods Alongside Gluten-Containing Food May Not Always Be as Risky for Celiac Patients as Diet Guides Suggest

INTRODUCTION

Celiac disease (CD) is a gluten-responsive inflammatory disorder with worldwide prevalence of 1%. Avoiding gluten, a protein found in wheat, rye, and barley, is inherently challenging because gluten is ubiquitous, tasteless, and not always visible. Many patients with CD eat GF food prepared in a home kitchen alongside gluten-containing foods. Fear of gluten exposure is common among CD patients and often leads to hypervigilance and decreased quality of life¹. The Codex Alimentarius Commission defines GF as <20 parts per million (ppm) gluten². GF diet guides published by professional societies, hospitals, and advocacy organizations suggest using dedicated toasters to prevent cross-contact with gluten and have conflicting recommendations regarding the need for dedicated utensils, leading to confusion for patients ^3,4. These recommendations are based upon expert consensus with scant data to support them⁵. The aim of this study was to quantify gluten transfer when GF foods are prepared alongside gluten-containing foods. A secondary aim was to assess the efficacy of cleaning methods for kitchen equipment/utensils.

Methods

Three scenarios were developed to assess gluten transfer and efficacy of washing methods during food preparation: (1) cooking pasta, (2) toasting bread, and (3) slicing cupcakes.

Cooking Pasta.

Sixteen-ounce packages of gluten-containing pasta (penne and fusilli) (Barilla USA) were boiled separately in stainless steel pots in fresh tap water for 12 minutes, then removed using handheld strainers. Water was reused to cook GF penne and fusilli (Dr.Schar USA). Cooked pasta was aliquoted into 2-ounce servings to quantify gluten transfer. The effect of rinsing was tested by running additional cooked and contaminated pasta under cold tap water for 30 seconds. To test washing methods, pots used for gluten-containing pasta were rinsed with water alone or scrubbed with soap and water, then used to cook GF pasta in fresh boiling tap water.

Toaster.

GF bread (Artisan White Bread, Dr Schar) was toasted in two rolling toasters (Spectrum, ECT-500) and (Belleco Conveyor Toaster, JT1) in a busy hospital cafeteria at 20-minute intervals, or in one of three shared pop-up toasters (Cuisinart, CPT-160, CPT-180, and Elite Cuisine, ECT-1027) immediately after toasting gluten-containing bread (Nature’s Own Whole 100% Wheat, Harvest Pride White and Wheat Sandwich Bread, Ultimate Grains Whole Grain Wheat Bread, or What’s A Bagel products). Gluten-containing crumbs were visible in all toasters. No cleaning procedures were attempted for toasters.

Slicing a Cupcake.

Frosted gluten-containing cupcakes (Vanilla Cupcakes, Whole Foods Market, Washington, DC) were sliced once with a knife, which was reused to slice a frosted GF cupcake (Vanilla Cupcakes, Whole Foods Gluten-Free Bake House). The knife was then washed in soap and water, rinsed in running water, or cleaned with an antibacterial hand wipe (Wet Ones) and a new GF cupcake was sliced. Both GF cupcakes were analyzed for gluten content.

Determination of Gluten Content and Data Analysis.

Samples were individually packaged in plastic bags with randomized sample numbers. Entire items (i.e., whole bread slice) were homogenized for analysis. Gluten content was assayed using an R5 sandwich ELISA (R7001, R-Biopharm, Darmstadt, Germany) which has a limit of detection of 5 ppm gluten by Bia Diagnostics (Colchester, Vermont). Control samples were also tested. Quantity of gluten for samples were categorized as <5ppm, 5-10ppm, 10-20ppm, or >20ppm. Confidence intervals are based upon binomial distribution.

RESULTS

Control samples of GF pasta, bread, and cupcakes all tested below the limit of detection. Gluten was detected in all pasta samples cooked in water used for gluten-containing pasta (33.9ppm to 115.7ppm). Rinsing pasta under running tap water reduced gluten content to less than 20 ppm. The two samples with detectable gluten had only 5.1 ppm and 17.5 ppm gluten. Rinsing pots with water alone after cooking gluten-containing pasta was as effective as scrubbing with soap and water to prevent detectable gluten transfer. Toasting in a shared toaster was not associated with gluten transfer above 20ppm; the four samples with detectable gluten had levels ranging only from 5.1 ppm to 8.3 ppm gluten. Although 28/30 cupcake samples had detectable gluten transfer, only 2/28 tested >20ppm. All three knife washing methods were effective in removing gluten.

DISCUSSION

Of the three scenarios tested, cooking GF pasta using shared water was the riskiest, resulting in gluten levels greater than 20 ppm in all samples tested. The risk of gluten cross-contact may be mitigated by rinsing the pasta, which reduced gluten content to less than 20 ppm. Alternatively, simply rinsing pots with clean water or scrubbing with soap and water before cooking GF pasta in fresh water may be a safer approach. Surprisingly, gluten transfer using a shared toaster was minimal, even with visible accumulation of gluten-containing crumbs in the toaster pans. Cutting cupcakes with a knife used to cut frosted gluten-containing cupcakes was associated with low-level gluten transfer even when crumbs were visible on the icing adhered to the knife.

Some kitchen activities may pose less of a risk of cross-contact with gluten than is commonly believed. Moreover, the risk may be readily mitigated by washing with water. In a small study, Studerus et al also found that dedicated kitchen equipment or preparation areas for GF foods may be unnecessary if appropriate cleaning methods are consistently employed⁵. In addition, Miller et al showed that a 2 meter radius from wheat flour and good hygiene practices facilitated simultaneous preparation of GF meals alongside gluten-containing ones⁶.

Acknowledged limitations of our study include small sample size, subsampling of homogenized foods, and use of sandwich rather than competitive ELISA which is able to detect hydrolyzed gluten.

Cross-contact will inevitably vary in different kitchen environments. It is important for CD patients to continue to ask questions about methods used to prepare GF foods, while at the same time recognizing that some practices may present a higher risk than others and vigilance must be directed appropriately. Future larger studies in home and commercial cooking environments should address a wider range of scenarios, cooking surfaces, and types of food. Such studies are needed to inform evidence-based recommendations for best practices for GF food preparation that balance risk of gluten exposure with harm from anxiety and hypervigilance.

Table 1:

Gluten content of GF pasta cooked in the same pot as gluten-containing pasta, GF bread toasted in toasters also used for gluten-containing bread and GF cupcakes cut with a knife used to cut gluten-containing cupcakes. Products containing <20ppm gluten are eligible for a GF label under the Codex Alimentarius Commission standard.

Preparation Method	Gluten Undetectable <5ppm		Gluten Detected 5ppm-10ppm		Gluten Detected 10ppm-20ppm		Gluten Detected >20ppm
	N (%)	95% CI	N (%)	95% CI	N (%)	95% CI	N (%)	95% CI
Gluten-Free Pasta
Cooked in shared water (n=12)	0	0-30%%	0	0-30%%	0	0-30%%	12 (100%)	70-100%
Cooked in shared water, then rinsed with cold water for 30 seconds (n=6)	4 (66%)	24-94%	1 (17%)	1-64%%	1 (17%)	1-64%	0	0-48%%
Shared pot washed with soap and water before cooking GF pasta (n=6)	6 (100%)	55-100%%	0	0-48%%	0	0-48%%	0	0-48%
Shared pot rinsed with water before cooking GF pasta (n=6)	6 (100%)	55-100%	0	0-48%	0	0-48%	0	0-48%
Toaster
GF bread toasted in communal rolling toaster (n=20)	16 (80%)	56-93%	4 (20%)	7-44%	0	0-20%	0	0-20%
GF bread toasted in shared Pop-Up Toaster (n=20)	20 (100%)	80-100%	0	0-20%	0	0-20%	0	0-20%
Cupcake
GF cupcake sliced with knife used to cut gluten-containing cupcake (n=30)	2 (7%)	1-24%%	12(40%)	23-59%	14 (46%)	29-65%	2 (7%)	1-24%
GF cupcake sliced with a washed knife (Table) N (%) Soap and Water Water Alone Wet Wipes	28 (93%) 9 9 10	76-99%	1(3%) 0 1 0	0.2-19%	1 (3%) 1 0 0	0.2-19%	0 (0%) 0 0 0	0-14%