Employee Dietary Initiative Improved Chronic Symptoms.

At the time of this health care improvement project, Essentia Health had nearly 15,000 employees, including 2100 physicians and advanced practitioners. It currently staffs 14 hospitals, 78 clinics, 6 long-term care facilities, 6 assisted living and independent living facilities, 7 ambulance services, 27 retail pharmacies, and 1 research institute. The sample included employees who took part in the Health Reset dietary challenge in 2021, 2022, and 2023. This health care improvement project was determined to not be research conducted with human subjects by Essentia Health’s human subjects’ protection program; thus, the health system’s institutional review board was not required to provide an exemption letter for this project. The Health Reset was offered to Essentia Health employees who took part in an employment wellness program annually in 2021, 2022, and 2023 as a challenge for which points could be accrued toward their health savings accounts. The Health Reset challenge was also mentioned in a system-wide weekly digest email to employees in each project year. The challenge aimed to report to participants the influence that their diets may have on their physical symptoms in 5 domains: energy level, sleep quality, gastrointestinal function, ability to concentrate, and aches/pains. The 28-day duration of the challenge was chosen so as to not far exceed the typical challenge length, generally up to 21 days. The half-life of food protein immunoglobulin G antibodies is around 3 weeks; 16 28 days can be sufficient for food elimination in many cases for which immunoglobulin G-mediated immune response is a primary driver of symptoms without causing nutritional imbalances. Participants were offered a physician-guided, 28-day nutritional challenge that removed dairy products, gluten-containing foods, and sugar from their diets in 2021, 2022, and 2023. An extensive guide (see Supplemental Material , available online at http://www.mcpiqojournal.org ) was provided, detailing the rationale for the intervention and foods to include and avoid. Recipes were also included. Weekly live question and answer sessions were provided before, during, and after each challenge period as foods were reintroduced. There were ∼6 sessions per challenge. Recordings of these sessions were made available to participants on the social media platform Yammer (now Viva Engage), hosted by the institution. Participants used this platform extensively to encourage one another, ask questions, and post recipes and other information. Survey questions (see Tables) were developed by a health care improvement project team member in collaboration with the physician lead. Only participants who completed both the pre-challenge and post-challenge surveys in a given year were included in analyses completed by the last author (MH). Demographic characteristics of Health Reset challenge participants were described using univariate statistics. The Wilcoxon signed rank test for matched pairs was used to compare presurvey and postsurvey responses to scale (1-10) items: “Please rate your current energy level” (energy level) (1=very low energy level, 10=very high energy level); ”Please rate any aches and pains you may currently be experiencing (including headaches)” (aches/pains) (1=no pain, 10=extreme pain); “Please rate any sleep difficulties you may be experiencing” (sleep difficulties) (1=no sleep trouble, 10=extreme sleep trouble); “Please rate any digestive issues you may be experiencing” (digestive issues) (1=no digestive issues, 10 =extreme digestive issues); “Please rate any difficulty concentrating you may be experiencing” (concentration difficulties) (1=no trouble concentrating, 10=extreme difficulty concentrating). Change variables were also created between presurvey and postsurvey responses to the energy level, aches and pains, sleep difficulties, digestive issues, and concentration difficulties questions, with all but change in energy level reverse coded to show improvements as positive values. Analysis of variance (ANOVA) and post hoc Tukey honestly significant difference tests were then used to compare mean change scores between grouped postsurvey diet adherence levels (0=did not/minimally adhered, 1=somewhat adhered, 2=mostly/completely adhered) for each project year separately. All change variables were normally distributed for each project year, with nonsignificant Levene’s test for equality of variances in each year, showing that variances were equal as required in ANOVA. Effect sizes were assessed using η 2 , where .01 represented a small, .06 a medium, and .14 a large effect. 17 Data were analyzed in IBM SPSS Statistics 29 (IBM Corp, 1998-2022). Missing data were excluded using listwise deletion. Comparisons were considered statistically significant at P< .05. We followed SQUIRE 2.0 guidelines in the preparation of this manuscript ( https://www.squire-statement.org ).

Over the 3 Health Reset dietary challenge years, 754 unique employees took the pre-challenge survey at least once, with 562 taking it in a single year, 149 taking it in 2 separate years, and 43 taking it all 3 years the challenge was offered. A total of 384 employees completed both the presurvey and postsurvey in at least 1 year. Regarding adherence to the Health Reset guidelines, most postsurvey participants in each year reported adhering somewhat (eliminated 1 or 2 of the 3 recommended foods or not consistently) or mostly (eliminated dairy, gluten, or sugar but had a couple slip-ups) to the Health Reset dietary guidelines ( Table 1 ), and most planned to maintain the changes they made to their diets during the challenge period. The most helpful supports offered during the Health Reset period included the Health Reset guide, followed by the Yammer group, and weekly support sessions (respondents could choose more than one). Table 1 Health Reset Dietary Challenge 2021-2023: Survey Completers and Postsurvey Categorical Questions 2021 2022 2023 n % n % n % Completed pre-challenge survey 282 100.0 356 100.0 351 100.0 Completed pre-challenge and post-challenge surveys 166 58.9% 187 52.5 153 43.6 Postsurvey: how closely did you adhere to the challenge guidelines?  Did not adhere (did not eliminate any of the recommended foods or decrease consumption) 3 1.8 3 1.6 4 2.6  Adhered minimally (did not eliminate any of the recommended foods but cut back on consumption overall) 13 7.8 18 9.6 14 9.2  Adhered somewhat (eliminated 1 or 2 of the 3 recommended foods or not consistently) 58 34.9 77 41.2 78 51.0  Adhered mostly (eliminated dairy, gluten or sugar but had a couple slip-ups) 74 44.6 71 38.0 49 32.0  Adhered completely (complete elimination of dairy, gluten, or sugar) 18 10.8 18 9.6 8 5.2 Postsurvey: what did you find most helpful about the challenge (select all that apply)  Challenge guide 131 78.9 152 81.3 132 86.3  Yammer group 118 71.1 109 58.3 51 33.3  Weekly support sessions 74 44.6 75 40.1 47 30.7  I did not find any of the challenge materials helpful 0 .0 4 2.1 3 2.0  Missing - - 1 .5 2 1.3 Postsurvey: Do you plan on maintaining any of the modifications you made throughout the challenge?  No 3 1.8 4 2.1 3 2.0  Yes 132 79.5 127 67.9 105 68.6  Maybe 31 18.7 55 29.4 44 28.8  Missing 0 .0 1 .5 1 .7 Participants who only completed the post-challenge survey (no matching pre-challenge survey) in a given year were excluded. Some participants also completed either the pre-challenge or post-challenge survey twice, and the duplicate survey was excluded. Health Reset Dietary Challenge 2021-2023: Survey Completers and Postsurvey Categorical Questions Participants who only completed the post-challenge survey (no matching pre-challenge survey) in a given year were excluded. Some participants also completed either the pre-challenge or post-challenge survey twice, and the duplicate survey was excluded. In each of the 3 dietary challenge years, participants who completed both pre-challenge and post-challenge surveys reported significantly greater mean ranked energy levels and lower mean ranked levels of aches/pain, sleep troubles, digestive issues, and difficulty concentrating after taking part in the dietary challenge (all P <.001; Table 2 ). Table 2 Health Reset Dietary Challenge 2021-2023: Differences Between Pre-Challenge and Post-Challenge Survey Continuous Measures a , b Items 2021 2022 2023 Pre-Challenge Post-Challenge Pre-Challenge Post-Challenge Pre-Challenge Post-Challenge n Mdn (IQR) N Mdn (IQR) P n Mdn (IQR) n Mdn (IQR) P n Mdn (IQR) n Mdn (IQR) P Energy level 166 6 (5-7) 165 8 (7-8) <.001 187 6 (5-7) 187 7 (6-8) <.001 153 6 (5-7) 153 7 (6.5-8) <.001 Pain level 165 5 (2.5-7) 165 2 (1-4) <.001 187 5 (3-6) 186 3 (2-5) <.001 153 5 (3-6) 151 3 (2-4) <.001 Sleep troubles 165 5 (2.5-6.5) 162 2.5 (1-4) <.001 185 5 (3-7) 186 2 (1-4) <.001 153 5 (3-6) 151 3 (2-4) <.001 Digestive issues 159 4 (2-6) 161 2 (1-3) <.001 185 5 (2-7) 184 2 (1-4) <.001 153 5 (3-7) 152 3 (1-4) <.001 Difficulty concentrating 165 4 (2-6) 164 2 (1-3) <.001 186 5 (3-6) 186 2 (2-4) <.001 153 5 (3-6) 153 3 (2-4) <.001 a Abbreviations: IQR, interquartile range; Mdn, median. b Wilcoxon signed rank test for matched pairs. Those who only completed the pre-challenge survey were excluded from the matched analyses. Each item is scored on a 1-10 scale, from lowest to highest. Health Reset Dietary Challenge 2021-2023: Differences Between Pre-Challenge and Post-Challenge Survey Continuous Measures a , b Abbreviations: IQR, interquartile range; Mdn, median. Wilcoxon signed rank test for matched pairs. Those who only completed the pre-challenge survey were excluded from the matched analyses. Each item is scored on a 1-10 scale, from lowest to highest. Descriptive statistics and ANOVA results for changes in scores between presurvey and postsurvey items by year and diet adherence level are presented in Table 3 . In 2021 ( Table 3 ), the first year of the dietary challenge, the change in energy level between presurvey and postsurvey differed significantly based on diet adherence level (F[2, 162]=8.02; P <.001), as did the change in digestive issues (F[2, 151]=4.84; P =.009). Effect sizes were small to medium in this and all project years, with energy level having the highest effect size (η 2 =.09, 95% CI, .02-.18) ( Table 4 ). Participants who mostly/completely adhered to the diet had significantly greater improvement in: energy levels (M=2.2, SD=1.8) versus those who did not/minimally (M=.8, SD=1.4, P =.007) and somewhat adhered (M=1.3, SD=1.7, P =.004); and digestive issues (M=2.4, SD=2.4) versus those who somewhat adhered (M=1.3, SD=2.0, P =.01) ( Table 5 ). Table 3 Health Reset Dietary Challenge Diet Adherence Descriptive Statistics and ANOVA for Postdietary Challenge Change Scores by Year and Dietary Challenge Adherence Level a Dependent Variable Challenge Diet Adherence n Mean SD SE ANOVA Comparison Sum of Squares df Mean Square F P 2021 Change in energy level Did not/minimally 16 .75 1.44 .36 Between groups 46.83 2 23.42 8.02 <.001 b Somewhat 57 1.25 1.70 .23 Within groups 472.78 162 2.92 Mostly/completely 92 2.17 1.75 .18 Total 519.61 164 Total 165 1.72 1.78 .14 Change in aches/pains c Did not/minimally 16 1.75 1.69 .42 Between groups 13.60 2 6.80 1.52 .22 Somewhat 57 1.28 2.02 .27 Within groups 720.62 161 4.48 Mostly/completely 91 1.90 2.24 .23 Total 734.22 163 Total 164 1.67 2.12 .17 Change in sleep difficulties c Did not/minimally 15 2.00 2.67 .69 Between groups 12.58 2 6.29 1.26 .29 Somewhat 56 1.34 1.91 .26 Within groups 791.84 158 5.01 Mostly/completely 90 1.91 2.35 .25 Total 804.42 160 Total 161 1.72 2.24 .18 Change in digestive issues c Did not/minimally 15 1.40 1.76 .46 Between groups 46.33 2 23.17 4.84 .009 b Somewhat 54 1.31 1.98 .27 Within groups 722.14 151 4.78 Mostly/completely 85 2.44 2.37 .26 Total 768.47 153 Total 154 1.94 2.24 .18 Change in concentration difficulties c Did not/minimally 16 1.50 1.93 .48 Between groups 4.51 2 2.25 .63 .54 Somewhat 58 1.60 1.84 .24 Within groups 575.16 160 3.59 Mostly/completely 89 1.91 1.92 .20 Total 579.67 162 Total 163 1.76 1.89 .15 2022 Change in energy level Did not/minimally 21 .62 1.75 .38 Between groups 58.95 2 29.47 12.14 <.001 b Somewhat 77 .86 1.46 .17 Within groups 446.83 184 2.43 Mostly/completely 89 1.92 1.60 .17 Total 505.78 186 Total 187 1.34 1.65 .12 Change in aches/pains c Did not/minimally 21 .33 1.71 .37 Between groups 58.50 2 29.25 7.33 .001 b Somewhat 77 1.06 1.91 .22 Within groups 730.06 183 3.99 Mostly/completely 88 1.94 2.13 .23 Total 788.56 185 Total 186 1.40 2.06 .15 Change in sleep difficulties c Did not/minimally 21 1.57 2.18 .48 Between groups 30.41 2 15.21 3.22 .04 b Somewhat 76 1.68 2.25 .26 Within groups 855.24 181 4.73 Mostly/completely 87 2.47 2.10 .23 Total 885.65 183 Total 184 2.04 2.20 .16 Change in digestive issues c Did not/minimally 20 .95 2.35 .53 Between groups 46.49 2 23.24 4.00 .02 b Somewhat 76 1.57 2.19 .25 Within groups 1046.10 180 5.81 Mostly/completely 87 2.38 2.60 .28 Total 1092.59 182 Total 183 1.89 2.45 .18 Change in concentration difficulties c Did not/minimally 21 .43 2.20 .48 Between groups 62.24 2 31.12 6.41 .002 b Somewhat 76 1.37 2.34 .27 Within groups 883.92 182 4.86 Mostly/completely 88 2.18 2.08 .22 Total 946.16 184 Total 185 1.65 2.27 .17 2023 Change in energy level Did not/minimally 18 .72 1.93 .46 Between groups 42.08 2 21.04 9.29 <.001 b Somewhat 78 1.13 1.44 .16 Within groups 339.70 150 2.26 Mostly/completely 57 2.11 1.44 .19 Total 381.78 152 Total 153 1.44 1.58 .13 Change in aches/pains c Did not/minimally 18 .83 1.76 .41 Between groups 10.36 2 5.18 1.36 .26 Somewhat 76 1.32 1.93 .22 Within groups 563.59 148 3.81 Mostly/completely 57 1.67 2.03 .27 Total 573.95 150 Total 151 1.39 1.96 .16 Change in sleep difficulties c Did not/minimally 18 1.00 1.57 .37 Between groups 28.46 2 14.23 3.56 .03 b Somewhat 76 1.16 2.04 .23 Within groups 591.09 148 3.99 Mostly/completely 57 2.02 2.06 .27 Total 619.55 150 Total 151 1.46 2.03 .17 Change in digestive issues c Did not/minimally 18 .72 2.16 .51 Between groups 46.16 2 23.08 4.35 .01 b Somewhat 77 1.52 2.30 .26 Within groups 790.34 149 5.30 Mostly/Completely 57 2.39 2.35 .31 Total 836.50 151 Total 152 1.75 2.35 .19 Change in concentration difficulties c Did not/minimally 18 .06 2.24 .53 Between groups 46.99 2 23.49 6.51 .002 b Somewhat 78 1.59 1.69 .19 Within groups 541.18 150 3.61 Mostly/completely 57 1.89 2.06 .27 Total 588.17 152 Total 153 1.52 1.97 .16 a Abbreviations: df, degrees of freedom; SD, standard deviation; SE, standard error. b Significant differences at P <.05. c Reverse coded. Table 4 ANOVA Effect Sizes by Health Reset Dietary Challenge Year a Dependent Variable η 2 95% Confidence Interval Lower Upper 2021 Change in energy level .090 .019 .175 Change in aches/pains b .019 .000 .070 Change in sleep difficulties b .016 .000 .064 Change in digestive issues b .060 .004 .138 Change in concentration difficulties b .008 .000 .046 2022 Change in energy level .117 .039 .201 Change in aches/pains b .074 .014 .149 Change in sleep difficulties b .034 .000 .094 Change in digestive issues b .043 .001 .106 Change in concentration difficulties b .066 .010 .139 2023 Change in energy level .110 .028 .202 Change in aches/pains b .018 .000 .071 Change in sleep difficulties b .046 .000 .118 Change in digestive issues b .055 .002 .132 Change in concentration difficulties b .080 .012 .165 a Abbreviation: η 2 =Eta-squared b Reverse-coded. Table 5 Tukey HSD Post Hoc Tests by Health Reset Dietary Challenge Year a Dependent Variable (I)Adherence (J) Adherence Mean Difference (I-J) SE P 95% Confidence Interval Lower Bound Upper Bound 2021 Change in energy level Did not/minimally Somewhat −.5 .48 .56 −1.64 .65 Mostly/completely −1.42 .46 .007 b −2.52 −.33 Somewhat Did not/minimally .5 .48 .56 −.65 1.64 Mostly/completely −.93 .29 .004 b −1.61 −.25 Mostly/completely Did not/minimally −1.42 .46 .007 b .33 2.52 Somewhat −.93 .29 .004 b .25 1.61 Change in aches/pains c Did not/minimally Somewhat .47 .6 .71 −.95 1.89 Mostly/completely −.15 .57 .96 −1.51 1.21 Somewhat Did not/minimally −.47 .6 .71 −1.89 .95 Mostly/completely −.62 .36 .19 −1.47 .22 Mostly/completely Did not/minimally .15 .57 .96 −1.21 1.51 Somewhat .62 .36 .19 −.22 1.47 Change in sleep difficulties c Did not/minimally Somewhat .66 .65 .57 −.88 2.2 Mostly/completely .9 .62 .99 −1.39 1.57 Somewhat Did not/minimally −.66 .65 .57 −2.2 .88 Mostly/completely −.57 .38 .29 −1.47 .33 Mostly/completely Did not/minimally −.09 .62 .99 −1.57 1.39 Somewhat .57 .38 .29 −.33 1.47 Change in digestive issues c Did not/minimally Somewhat .09 64 .99 −1.43 1.60 Mostly/completely −1.04 .61 .21 −2.48 .41 Somewhat Did not/minimally −.09 .64 .99 −1.60 1.43 Mostly/completely −1.12 .38 .01 b −2.02 −.22 Mostly/completely Did not/minimally 1.04 .61 .21 −.41 2.48 Somewhat −1.12 .38 .01 b .22 2.02 Change in concentration difficulties c Did not/minimally Somewhat −.10 .54 .98 −1.37 1.16 Mostly/completely −.41 .51 .71 −1.63 .81 Somewhat Did not/minimally .10 .54 .98 −1.16 1.37 Mostly/completely −.31 .32 .60 −1.06 .45 Mostly/completely Did not/minimally .41 .51 .71 −.81 1.63 Somewhat .31 .32 .60 −.45 1.06 2022 Change in energy level Did not/minimally Somewhat −.24 .38 .81 −1.14 .67 Mostly/completely −1.30 .38 .002 b −2.20 −.41 Somewhat Did not/minimally .24 .38 .81 −.67 1.14 Mostly/completely −1.06 .24 <.001 b −1.64 −.49 Mostly/completely Did not/minimally −1.30 .38 .002 b .41 2.20 Somewhat −1.06 .24 <.001 b .49 1.64 Change in aches/pains c Did not/minimally Somewhat −.73 .49 .30 −1.89 .43 Mostly/completely −1.61 .49 .003 b −2.76 −.46 Somewhat Did not/minimally .73 .49 .30 −.43 1.89 Mostly/completely −.88 .31 .01 b −1.61 −.14 Mostly/completely Did not/minimally −1.61 .49 .003 b .46 2.76 Somewhat −.88 .31 .01 b .14 1.61 Change in sleep difficulties c Did not/minimally Somewhat −.11 .54 .98 −1.38 1.15 Mostly/completely −.90 .53 .21 −2.15 .35 Somewhat Did not/minimally .11 .54 .98 −1.15 1.38 Mostly/completely −.79 .34 .06 −1.59 .02 Mostly/completely Did not/minimally .90 .53 .21 −.35 2.15 Somewhat .79 .34 .06 −.02 1.59 Change in digestive issues c Did not/minimally Somewhat −.62 .61 .57 −2.05 .82 Mostly/completely −1.43 .60 .047 b −2.84 −.02 Somewhat Did not/minimally .62 .61 .57 −.82 2.05 Mostly/completely −.81 .38 .08 −1.71 .08 Mostly/completely Did not/minimally −1.43 .60 .047 b .02 2.84 Somewhat .81 .38 .08 −.08 1.71 Change in concentration difficulties c Did not/minimally Somewhat −.94 .54 .20 −2.22 .34 Mostly/completely −1.75 .54 .004 b −3.02 −.49 Somewhat Did not/minimally .94 .54 .20 −.34 2.22 Mostly/completely −.81 .35 .05 −1.63 .00 Mostly/completely Did not/minimally −1.75 .54 .004 b .49 3.02 Somewhat .81 .35 .05 .00 1.63 2023 Change in energy level Did not/minimally Somewhat −.41 .39 .56 −1.34 .53 Mostly/completely −1.38 .41 .002 b −2.35 −.42 Somewhat Did not/minimally .41 .39 .56 −.53 1.34 Mostly/completely −.98 .26 <.001 b −1.60 −.36 Mostly/completely Did not/minimally −1.38 .41 .002 b .42 2.35 Somewhat −.98 .26 <.001 b .36 1.60 Change in aches/pains c Did not/minimally Somewhat −.48 .51 .61 −1.69 .73 Mostly/completely −.83 .53 .26 −2.08 .42 Somewhat Did not/minimally .48 .51 .61 −.73 1.69 Mostly/completely −−.35 .34 .56 −1.16 .46 Mostly/completely Did not/minimally .83 .53 .26 −.42 2.08 Somewhat .35 .34 .56 −.46 1.16 Change in sleep difficulties c Did not/minimally Somewhat −.16 .52 .95 −1.40 1.08 Mostly/completely −1.02 .54 .15 −2.30 .26 Somewhat Did not/minimally .16 .52 .95 −1.08 1.40 Mostly/completely −.86 .35 .04 b −1.69 −.03 Mostly/completely Did not/minimally 1.02 .54 .15 −0.26 2.30 Somewhat −.86 .35 .04 b .03 1.69 Change in digestive issues c Did not/minimally Somewhat −.80 .60 .39 −2.22 .63 Mostly/completely −1.67 .62 .02 b −3.14 −.19 Somewhat Did not/minimally .80 .60 .39 −.63 2.22 Mostly/completely −.87 .40 .08 −1.82 .09 Mostly/completely Did not/minimally −1.67 .62 .02 b .19 3.14 Somewhat .87 .40 .08 −.09 1.82 Change in concentration difficulties c Did not/minimally Somewhat −1.53 .50 .007 b −2.71 −.36 Mostly/completely −1.84 .51 .001 b −3.05 −.62 Somewhat Did not/minimally −1.53 .50 .007 b .36 2.71 Mostly/completely −.30 .33 .63 −1.09 .48 Mostly/completely Did not/minimally −1.84 .51 .001 b .62 3.05 Somewhat .30 .33 .63 −.48 1.09 a Abbreviation: HSD, honestly significant difference; SE, standard error. b Significant difference at P< .05. c Reverse coded. Health Reset Dietary Challenge Diet Adherence Descriptive Statistics and ANOVA for Postdietary Challenge Change Scores by Year and Dietary Challenge Adherence Level a Abbreviations: df, degrees of freedom; SD, standard deviation; SE, standard error. Significant differences at P <.05. Reverse coded. ANOVA Effect Sizes by Health Reset Dietary Challenge Year a Abbreviation: η 2 =Eta-squared Reverse-coded. Tukey HSD Post Hoc Tests by Health Reset Dietary Challenge Year a Abbreviation: HSD, honestly significant difference; SE, standard error. Significant difference at P< .05. Reverse coded. In 2022 ( Table 3 ), all change variables differed significantly based on adherence level: energy level (F[2, 184]=12.14; P <.001), aches and pains (F[2, 183]=7.33, P =.001), sleep difficulties (F[2, 181]=3.22; P =.04), digestive issues (F[2, 180]=4.00; P =.02), and concentration difficulties (F[2, 182]=6.41; P =.002). The change in energy level again had the highest effect size (η 2 =.12, 95% CI, .04-.20) ( Table 4 ). Participants who mostly/completely adhered to the diet had significantly greater improvement in: energy levels (M=1.9, SD=1.6) versus those who did not/minimally (M=.6, SD=1.8, P =.002) and somewhat adhered (M=.9, SD=1.5, P <.001); aches/pains (M=1.9, SD=2.1) versus did not/minimally (M=.3, SD=1.7, P =.003) and somewhat adhered (M=1.1, SD=1.9, P =.01); digestive issues (M=2.4, SD=2.6) versus those who did not/minimally adhered (M=1.0, SD=2.4, P =.047); and difficulty concentrating (M=2.2, SD=2.1) versus those who did not/minimal adhered (M=.4, SD=2.2, P =.004) ( Table 5 ). In 2023 ( Table 3 ), aside from aches/pains, 4 change variables differed significantly based on adherence level: energy level (F[2, 150]=9.29, P <.001), sleep difficulties (F[2, 148]=3.56, P =.03), digestive issues (F[2, 149]=4.35, P =.02), and concentration difficulties (F[2, 150]=6.51, P =.002). As in previous years, energy level had the highest effect size (η 2 =.11, 95% CI, .03-.20) ( Table 4 ). Participants who mostly/completely adhered to the diet had significantly greater improvement in: energy levels (M=2.1, SD=1.4) versus those who did not/minimally (M=.7, SD=1.9, P =.002) or somewhat adhered (M=1.1, SD=1.4, P <.001); sleep difficulties (M=2.0, SD=2.1) versus those who somewhat adhered (M=1.2, SD=2.0, P =.04); digestive issues (M=2.4, SD=2.4) versus those who did not/minimally adhered (M=.7, SD=2.2, P =.02); and concentration difficulties (M=1.9, SD=2.1) versus those who did not/minimally adhered (M=.6, SD=2.2, P =.001) ( Table 5 ). Participants who somewhat adhered to the diet also had greater improvement in concentration difficulties (M=1.6, SD=1.7) versus those who did not/minimally adhere ( P =.007).

Essentia Health, a self-insured integrated rural health delivery system headquartered in Duluth, MN and serving patients in Minnesota, Wisconsin, and North Dakota, implemented an employee-based initiative called the Health Reset to evaluate and improve employee health from 2021 to 2023. The aim of this health care improvement project was to evaluate the Health Reset dietary challenge’s effect on participants’ self-reported energy levels, sleep quality, gastrointestinal function, ability to concentrate, and aches/pains in each of the 3 project years.

Essentia Health’s Health Reset health care improvement project appeared to improve self-reported outcomes, increase engagement with employees, and has the potential to reduce long-term health care costs if the diet is maintained to keep symptoms under control. Energy levels and digestive symptoms improved consistently with greater adherence to the challenge, suggesting improved vitality from a relatively simple initiative outside of clinical care. Suggested next steps include formal research, which could validate this instrument, confirm our findings in other populations and in randomized control trials, and explore the benefits of voluntary health challenges to groups of employees and patients in general.

Standard American Diet has been shown to be a contributor to many common chronic diseases. 1 , 2 , 4 , 18 Although controversy exists around which foods may be detrimental to health, consuming a diet that is nutrient dense reduces all-cause mortality. 19 In this health care improvement project evaluation of a physician-guided, 28-day Health Reset dietary challenge—removing dairy products, gluten-containing foods, and sugar—offered yearly from 2021 to 2023 to health system employee participants, we found significantly improved mean ranked scores for self-reported energy levels, aches/pains, sleep difficulties, digestive issues, and difficulty concentrating between predietary and postdietary challenge surveys. Furthermore, changes in all items measured were statistically significant when comparing adherence levels in at least 1 project year. Participants who most closely adhered to the diet had significantly higher improvements in self-reported energy levels and digestive issues in all project years. This suggests that those who most closely adhered to the diet restrictions may have experienced the most perceived benefit in these 2 areas at the end of the challenge. The detrimental effects of sugar are understood. A hallmark of the Western diet, high sugar intake is well recognized as a risk factor for obesity, cardiovascular disease, metabolic syndrome, and type 2 diabetes. Excessive intake of dietary sugars can upregulate the production of pro-inflammatory cytokines and lead to chronic inflammation. 13 The consumption of sugar sweetened beverages has significantly contributed to the burden of cardiometabolic disease as the largest source of added sugar in the SAD, although its use is declining in some countries, including the United States, because of awareness campaigns and taxation. 20 Although micronutrients are bioavailable in 100% fruit juice, it is increasingly discouraged by health agencies because of its high glycemic load and a higher all-cause mortality risk associated with each serving of fruit juice. 21 Whole fruits contain numerous beneficial phytonutrients, less juice per serving, and fiber and are considered high quality food options. 22 , 23 Although the literature is sparse relative to gluten’s association with chronic disease and inflammation and this is currently a topic of debate, 24 , 25 , 26 some patients without celiac disease who have gastrointestinal or extra-intestinal symptoms can find improvement when they eradicate gluten from their diets. 27 Conditions associated with this so-called nonceliac gluten sensitivity include rheumatoid, osteoarthritis, autism spectrum disorder, endometriosis, nonspecific joint pain, fibromyalgia, fatigue, and autoimmunity. 14 , 28 , 29 , 30 , 31 , 32 Dairy products (eg, cow, sheep, goat, and other mammal milk) have also been associated with irritable bowel syndrome and constipation, 33 , 34 , 35 , 36 rashes, joint inflammation, 37 and respiratory or allergic symptoms 38 in susceptible individuals. Such symptoms appear to be mediated by lactose or casein (possibly primarily from A1 protein). 39 Despite the paucity of evidence for dairy as a cause of chronic symptoms, integrative medicine clinicians may recommend trials of dairy abstinence. Integrative clinicians commonly find that patients who shift from the SAD to some form of an anti-inflammatory diet typically note improvements in their well-being, an observation that is supported by peer reviewed literature studying the effects of an anti-inflammatory diet for common conditions. 9 , 40 , 41 As adherence to dietary recommendations is a critical factor for preventing and managing chronic diseases, it is important to identify and address barriers to adherence. 42 , 43 The productivity losses from chronic disease among employees are substantial; 44 workplace health promotion programs can favorably deliver on employee health and well-being and may be particularly appealing for self-insured companies. 45 , 46 , 47 The Health Reset challenge was voluntary; it is possible that more engaged employees already planning to modify their lifestyles were self-selected. Some employees may also not have been aware of the Health Reset challenge if they did not read the weekly digest email that mentioned the challenge or were not active in the employment wellness program at the time of each annual challenge. The noncompletion rate was high in each challenge year, ranging from 41% to 56%. The low participation numbers as a percentage of the employee population may be related to the highly challenging nature of the initiative. Future research could examine methods to increase participation, decrease attrition, and increase dietary challenge adherence, including follow-up with those who dropped out of similar challenges to understand barriers to completing the challenge. We also lacked evidence to support findings that partial dietary restriction adherence improved some health outcomes, which is another area for future research. Moreover, we lacked data on the number of celiac or estimated gluten sensitive nonceliac challenge participants. We also lacked respondents’ baseline health and body composition data, and we did not assess baseline dietary patterns, total calories or amount and types of fat consumed, fruit or vegetable intake, level of physical activity, frequency of eating out or eating fast foods, or other factors or lifestyle changes that may influence overall diet and health. Future research could assess the impacts of these factors on similar dietary challenge outcomes. It is worth noting that foods marketed as sugar-free, dairy-free, or gluten-free are not necessarily wholesome products; and in fact, some are highly processed and low in fiber. Participants were encouraged to consume a whole foods diet high in plants in the challenge guide and during the weekly question and answer sessions. Though we recommended avoiding processed foods, we did not inquire about this in the post-challenge questions. We acknowledge that removing whole grains from the diets of patients who do not react unfavorably to them could have detrimental effects if the replacement foods are refined or the eradication is unnecessarily prolonged. There was no control group or matched cohort, all responses were self-reported, and there was no assessment beyond the dietary challenge period. We also did not track which foods (sugar, dairy, or gluten) seemed to trigger symptoms when participants systematically reintroduced them, so these results reflect abstinence from all 3 foods together to the degree that participants adhered to the challenge guidelines. We were unable to assess whether positive outcomes were due to dietary changes or possibly reduced caloric intake. Finally, gastrointestinal symptoms can be wide-ranging, and the challenge questionnaires did not require participants to specify their symptoms, which makes understanding the effects of the Health Reset in this area more difficult. To mitigate some limitations of studying this initiative, we did control for the self-reported level of adherence to the dietary challenge in each of the 3 health care improvement project years.

The authors report no competing interests.