EFFECT OF HIGH- AND LOW-FODMAP DIET INSTRUCTION ON FODMAP INTAKE AND DIETARY QUALITY IN HEALTHY YOUNG ADULTS

Objective: Irritable Bowel Syndrome (IBS) is the most commonly diagnosed Gastrointestinal (GI) condition in the United States affecting 30 million (10%) people. Symptoms include abdominal pain, bloating, distension, excessive wind and altered bowel habits when anatomical abnormalities and inflammation have been excluded. A low-FODMAP diet is now considered as an effective strategy for managing symptoms of IBS in Australia, with interest expanding across the world. Several limitations of a lowFODMAP diet pertaining to dietary quality and health benefits have been suggested. Malabsorbed FODMAPs provide multiple benefits which include a natural laxative effect due to their osmotic effects, a prebiotic effect with beneficial fermentation by-products and production of a low glycemic response compared to other carbohydrates. Additionally, Dietary adherence is crucial to the success of a low-FODMAP diet, however most people do not find the diet easy to incorporate into their life. Not one study has looked exclusively at healthy adults or looked at changes in FODMAP intake and diet quality as compared to established guidelines. A study is needed looking at dietary quality of lowvs. high-FODMAP diets and should consider how adherence and other factors that may influence efficacy of the diet. Design: This study used a single-blinded crossover design. Subjects (n=16) were instructed about following a low-FODMAP and a high-FODMAP diet for three days each, presented in a random order and separated by an 11-day wash out period. The study was entitled “The Carb Study” and diets labeled “diet 1” and “diet 2” without reference to FODMAP. No food was provided. Dietary instruction was provided for each diet along with a dietary booklet. Dietary assessment consisted of four 24-hour recalls using NDSR. Recalls assessed the day prior to each intervention period (2 baselines) and assessed day 3 of each intervention period (2 interventions). FODMAP intake was estimated based on the sum of fructose, lactose and polyol intake and dietary quality was calculated based on the Healthy Eating Index 2010 (HEI-2010). Setting: Free living subjects recruited from a northeastern university. Subjects: Participants were healthy adults without gastrointestinal disorders (n=16, 63% female, 20.47±1.77 years). Results: There was no effect of diet order. There was a non-significant trend for a between treatment difference in FODMAP intake (F(1,14df)=4.27, p=.058) and a significant difference between groups in HEI-2010 total score (F(1,14df)=10.45, p=.001). Within the low-FODMAP treatment, FODMAP intake decreased from 36.30±22.62 grams to 19.29±15.79 grams (-t=2.84, p=.01) and HEI-2010 scores increased from 53.60±17.16 to 63.09±17.23 (t=2.20, p=.04); Energy intake also significantly decreased from 2259±1325 kcals to 1510±795 kcals (-t=2.68, p=.017). Within the high-FODMAP treatment, there was no change in FODMAP intake (t=.35, p=.731) but HEI-2010 scores decreased from 60.83±12.76 to 52.04±11.27 (-t=2.45, p=.027); There was no difference in energy intake (1993±962 to 2251± 864, -t=1.57, p=.14) Conclusions: This study suggests that reducing FODMAP is feasible in healthy, freeliving young adults and that this reduction is associated with an increase in dietary quality. However, the high-FODMAP intervention in this study was not effective in increasing FODMAP intake. Future research with larger samples is needed to develop interventions for increasing healthy FODMAP intake in young adults. In addition, future research is needed to assess long-term effects of these dietary modifications in healthy

small and osmotically-active, increasing the liquidity of luminal content due to osmosis and 3) rapidly fermented by gut microbiota, increasing the amount of gas in the colon 1 .
These characteristics combine to increase luminal distension, the physiological basis for the genesis of many GI symptoms 1 .
It is hypothesized that GI symptoms are created primarily by luminal distention 3 creased by fermentation and osmosis 4 . Studies have concluded that high-FODMAP diets induce GI symptoms 5,6 , and low-FODMAP diets relieve GI symptoms associated with functional GI disorders 7-10 with GI symptoms returning when FODMAPs are reintroduced into the diet 11 . Overall GI symptoms have been seen in up to 86% of Irritable Bowel Syndrome (IBS) patients 7 . Accordingly, a low-FODMAP diet has been recommended for managing GI symptoms for IBS patients 12 . Applications are expanding to enteral feeding formulas 13 , a low-FODMAP diet for patients with non-celiac gluten sensitivity 14 and the treatment of infantile colic 15 .
A first consideration for FODMAP diets is the overall quality and adequacy of the diet, as well as associated health benefits. FODMAPs are CHO or related polyols found in fruits, vegetables, legumes, wheat and other grain products as well as milk and dairy products. A FODMAP restricted diet limits the available options in these nutrientdense food groups. Additionally, in all populations FODMAPs are malabsorbed [16][17][18][19][20] .
Malabsorbed CHO can provide a prebiotic effect due to fermentation by-products 12 and CHO products high in FODMAPs tend to generate a lower glycemic response compared to CHO products lower in FODMAPs 21 . A low-FODMAP diet may adversely affect gut microflora and compromise fiber intake 12 and dietitians instructing patients on low-FODMAP diet should provide options for high-fiber alternative fruits, vegetables and grains as well as adequate sources of calcium and vitamin D 4,16,22 . There is limited evidence comparing the effect of low-vs. high-FODMAP diets on fiber or any other nutrient intake 12 . One retrospective study found limited differences in macronutrient intake comparing current diets of free-living subjects who had received low-FODMAP dietary advice two years previously to healthy controls 23 .
Another consideration for FODMAP diets is dietary adherence. Adherence appears to be crucial to the success of a low-FODMAP diet with correlations between adherence and symptom improvement reported 8 . Most people do not find the diet easy to incorporate into their life 8,10 , although controlled studies with IBS patients have shown high adherence rates both when all foods are provided 11 and when provided with dietary advice 8 . Potential barriers to adherence include buying the appropriate food 10 , implementing the diet 8,10 , following the diet 8 and taste 8,10 . There is limited research available about following a high-FODMAP diet, but individuals are likely to face barriers given the presence of adverse GI symptoms 5,6 .
The primary aim of this study was to investigate the effect of dietary instruction on implementing low-and high-FODMAP diets on FODMAP intake as well as on dietary quality in healthy, free living, young adults. Secondary aims were to investigate changes in mood, GI symptoms and subjects' opinions regarding the diet, as potential variables that may impact dietary adherence.

Subjects
Twenty healthy, young adults, free from GI disorders were enrolled in this study. Four subjects withdrew from this study, one for medical conditions unrelated to the study and three failed to compete any assessments beyond baseline. Thus 16 subjects completed the study and were considered the study sample. Exclusionary GI disorders included celiac disease, IBS, lactose or gluten intolerance, diverticular disease, colitis such as Crohn's disease or ulcerative colitis and stomach ulcers. Additional exclusion criteria included currently following a weight loss diet, food allergies, smoking, pregnancy or lactation, diabetes, adrenal disease, kidney or bladder problems, a thyroid disease or currently taking any appetite suppressant medication. All subjects were recruited via classroom announcements at the University of Rhode Island or emails sent to adults who were candidates/participants in previous, nutrition-related studies. Subjects received a $80 stipend for completing the study. The study was approved by the Institutional Review Board of the University of Rhode Island and subjects provided written informed consent prior to participating.

Study Design
The study was a randomized, single-blinded, cross-over study that compared two diet-interventions in a free-living setting; instruction on low-vs. high-FODMAP diet. In order to ensure that the diets were single-blinded, the study was entitled "The Carb Study" and the two interventions were labeled "Diet 1" and "Diet 2" representing the low-FODMAP diet and high-FODMAP diet respectively. No food was provided. Each intervention had a corresponding diet instruction booklet that was developed specifically for this project with foods identified as either high-or low in FODMAPs at the time of the study 1,5,7,[23][24][25] . Subjects were provided with 15-minute instructions about each dietary treatment and asked to follow this booklet to the best of their ability for each three day intervention period.
An initial screening was conducted to verify potential participants met inclusion criteria. Body fat percentage was assessed using the BOD POD Body Composition System (Life Measurement Instruments, Concord, Calif., USA). Subjects where then randomized to start with either the low-FODMAP diet (order 1) or the high-FODMAP diet (order 2). Each dietary period was followed by an eleven day wash out period where subjects consumed their normal diet. Following the wash out period, subjects completed the remaining dietary intervention. Each intervention lasted three days: Tuesday, Wednesday, and Thursday.
Four 24 hour recalls were conducted with each subject. At the start of each dietary intervention period on Tuesday, each 24-hour recall measured dietary intake on Monday (baseline). After completion of each intervention period on Friday, each recall measured intake on Thursday (intervention). FODMAP intake, Healthy Eating Index-2010 (HEI-2010) scores, and intake of other selected nutrients were obtained from these 24 hour recalls. Subjects also completed questionnaires on Fridays of each intervention period assessing mood and GI symptoms. An additional "opinion regarding the diet" questionnaire was filled out by a convenience sample of subjects at the end of each intervention period.

Dietary Instructions and Diet Booklet
At the baseline visit for each diet, subjects met with a trained research assistant who provided each subject with a 16 page diet instruction booklet. The booklets contained a detailed list of recommended and restricted foods corresponding to either the low-FODMAP diet or the high-FODMAP diet. A brief, 15 minute diet instruction session was provided, which included identifying encouraged and discouraged foods, tips, and emphasized the importance of adhering to the diet for research purposes.

NDSR 24 Hour Food Recall
As described above, four in-person 24 hour recalls were conducted by trained research assistants. Recalls were conducted using the Nutrition Data System for Research (NDS-R) software version 12 developed by the Nutrition Coordinating Center (NCC), University of Minnesota, Minneapolis. NDS-R utilizes a multiple pass method described in full elsewhere 26 . Briefly, pass one included obtaining a quick list of foods consumed in the past 24 hours. In pass two, participants were asked to produce details regarding foods on their quick list including portion sizes and amounts eaten. In pass three the list was recited and participants are asked if any information was forgotten. Food amount booklets distributed by the NCC as well as food models were available during food recalls in order to assure accurate portion sizes.
Foods that were not in the NDS-R database were listed as "missing foods" and resolved after the interview was completed. Resolution of a missing food required finding an NDS-R substitute (similar food or a generic version of a food) in the database and matching that substitute for CHO, protein, fat and kcal. Matching was defined as within 1-3 grams for each macronutrient and within 10 kcals for energy. For some foods, the potential FODMAP content was considered too variable for application of the normal missing food substitution protocol (for example ice cream brands and artificially sweetened beverages). These foods were sent to the NCC, who provided an accurate nutrient breakdown for those items.

FODMAP intake
NDS-R output files were used to sum total intake (g) of fructose, lactose and the polyols (erythritol, inositol, isomalt, lactitol, maltitol, mannitol, pinitol, sorbitol and xylitol  29 . The total HEI score ranges from 0 (low) to 100 (high). The total score is based on eight "adequacy" components: total fruit, whole fruit, total vegetables, greens and beans, whole grains, dairy, total protein foods, seafood and plant proteins as well as four "moderation" components: fatty acids, refined grains, sodium, and empty calories 29 . Intake is in energy adjusted per 1,000 kcal.
Calculation of the HEI-2010 for this study was based on a protocol developed by the NCC based on methods described in a previous study 30 . The calculations were made using Microsoft Excel, 2007 and performed twice to check accuracy.

Assessment of mood state and gastrointestinal symptoms
All subjects completed two questionnaires regarding their mood state and GI symptoms prior to bed on the final day of each intervention. The mood questionnaire used 10-cm VAS response scale anchored at each end "0=Very Little" and "10=Very Much" that had been used in a previous study 31

Opinion regarding the diet
Following each intervention 24 hour recall, a convenience sample completed a 4item evaluation of the diet using with a 10-cm VAS response scale from previous FODMAP-related studies 8,10,11 . Items included: how easy/difficult had it been to implement the diet, how easy/difficult had it been to adhere to the diet, how easy/difficult was it to obtain the appropriate food, how would you rank the overall taste.

Statistics
Statistical analyses were performed using SPSS, version 22.0 (IBM Corporation, Summers, NY, USA). All variables met criteria for normality. Baseline comparisons between subjects assigned to the two orders were conducted using t-tests and χ 2 tests.
Energy intake (kcal/day) was assessed using similar analyses. All other inferential analyses of dietary components were performed using a 2 x2 repeated measures ANOVA where the independent variables were treatment (low-vs. high-FODMAP) and time (baseline vs. intervention). Paired t-tests compared mood, symptoms and compliance factors between treatments and Pearsons bivariate correlations explored relationships between variables. Median scores were reported for diet opinion. P < .05 was considered statistically significant.

RESULTS
There were no differences in demographic variables between subjects assigned to the two orders. Mean age of the 16 subjects was 20.6 years (range 18-23) and 10 were female. Further information is presented in Table 1.
Values for FODMAP intake and all dietary data including both the total and subscales of the HEI-2010 did not differ at baseline between the two orders.
There were non-significant trends for grams of FODMAP per day for time

Summary
The purpose of this study was to investigate the effects of dietary instruction for low-and high-FODMAP diets on dietary quality and FODMAP intake in healthy young adults. We found that the low-FODMAP diet resulted in an increase in dietary quality with a reduction in FODMAP intake. The high-FODMAP diet had no effect on

Subjects
Healthy young adults free of any GI disorders were included in this study. To the researchers' knowledge, this is the only FODMAP study that has looked exclusively at healthy adults. Previous FODMAP focused on populations with IBS [5][6][7][8][9][10] . In one of the few FODMAP studies that included healthy adults, Ong et al. 5 compared healthy adults (n=15) to IBS patients (n=15) and compared high-FODMAP conditions (50g/day) to low-FODMAP conditions (9 g/day) during a two day intervention. The study was a singleblinded crossover intervention comparing low-vs. high-FODMAP conditions in which all foods were provided for two days. The study found that a high-FODMAP diet had no effect on symptoms except for increased flatulence in healthy adults. Similarly, our study found that both abdominal pain and distention were not factors in this population however we did not assess flatulence.

The effect of a low-FOMAP diet on Dietary Quality
Due to the restriction of fruits, vegetables, dairy and legumes, dietary quality was hypothesized to decrease on the low-FODMAP diet, whereas the opposite occurred.
Looking at the change in dietary quality, the low-FODMAP intervention was most effective at restricting refined products and "empty calories" including solid fat, added sugar and sodium-rich foods. This restriction contributed to the large decrease in energy and carbohydrate intake. These results indicate that a low-FODMAP diet has potential to have a positive influence on dietary quality in college students but future studies are needed with adults showing more dietary diversity. The implications of the decrease in energy intake should be examined in future research.
To the researchers' knowledge this was the first FODMAP study that examined dietary quality in healthy adults. Ostgaard et al. 23 examined diet composition of IBS patients who received low-FODMAP education (guided n=43) two years prior. This study showed no difference between the guided and control group for calories, CHO, protein, fat, sugars or fiber intake and did not measure dietary quality. This study used food frequency questionnaires (FFQ) to measure nutrient intakes as opposed to our study which used 24 hour food recalls.

The effect of a low-FODMAP diet on other health benefits
Fibers provide plant structure and are thus found in plant-derived vegetables, fruits, whole grains and legumes 35 . A low-FODMAP diet restricts these food groups, suggesting fiber intake might be of concern. In our study however fiber intake did not change in either intervention. Total, soluble and insoluble fiber intake were considerably lower than the recommendations 36

Strengths and Limitations
There were several limitations. FOS and GOS were not included because they are not analyzed by NDS-R. Future studies could use Barrett and Gibson's food frequency questionnaire 46 . A second limitation was that only a single 24 hour food recall was used per intervention period which provides an unstable estimate of usual intake 26,47 .
A third limitation was that the high-FODMAP intervention was not effective at reducing FODMAP intake. Other limitations to this study include a small sample size, a short duration and reduced generalizability due to the homogeneity of college-aged subjects.
Strengths of the study included the use of the HEI-2010, a well controlled, randomized single-blinded crossover experimental design, and use of healthy adults.

CONCLUSION
Dietary instruction for implementing a low-FODMAP diet may be effective in helping young healthy individuals reduce FODMAP intake without compromising overall dietary quality. Although calcium intake was low, this study found that the low-FODMAP diet was associated with a reduction in overall energy and carbohydrate intake as well as glycemic load. Long term studies are needed to confirm these results. Future research is also needed to assess the effects of increasing FODMAP intake in young adults.

21.
Liljeberg HGM, Akerberg AKE, Bjorck IME. Effect of the glycemic index and content of indigestible carbohydrates of cereal-based breakfast meals on glucose tolerance at lunch in healthy subjects.    A 2 (Order) X 2 (Treatment) X 2 (time) mixed factorial ANOVA with post hoc ttest was used for total HEI-2010 and a 2 (treatment) X 2 (Time) repeated measured ANOVA with post hoc ttest was used for individual components a = Time*treatment interaction F statistic reported b = A measure of dietary quality reflecting federal guidelines. Scores range from 0-100 with higher scores reflecting better diet quality. The value is expressed as a per 1,000 kcal standard. IBS's impact on the healthcare system has also been heavily researched. The overall associated cost is 1.6 billion in direct and 19.2 billion in indirect annual costs 56 .
The mean annual direct health care cost per patient is $5,049 and the annual out-ofpocket expenses (for example non-prescription medication and alternative treatment like special diets and therapy per patient) is $406 57 . The individual cost has been found to increase based on disease severity and recent exacerbation of bowel symptoms 57 .
Regarding burden for healthcare practitioners, IBS accounts for 12% of the patients seen in the primary care practice and is the largest diagnostic group seen in GI practice 58 with inpatient care accounting for 17.5% of total costs 57 .

Treatment of IBS Focusing on the Role of Diet
Current treatments for IBS include pharmaceuticals such as antispasmodics and stool softeners 12 , psychological therapy, fiber, probiotics and lifestyle and diet

FODMAP: Definition & General Properties
In 2005 the term FODMAP was coined to identify a collection of poorly digested, highly osmotic and rapidly fermented short chain carbohydrates (CHO). FODMAPs have three common functional properties. They are 1) poorly absorbed in the proximal small intestine, allowing substrate to reach the distal small intestine and proximal colon 2) small and osmotically-active, which increases the liquidity of luminal content due to osmosis and 3) rapidly fermented by gut microbiota, increasing the amount of gas present in the colon. These three characteristics combined to increase luminal distension 1 , the basis for the genesis of many functional gut syndrome 1 .
A low-FODMAP diet is now considered an effective strategy for managing symptoms of IBS in Australia, with interest expanding across the world 12  Improved symptoms included reduced bloating, abdominal pain and flatulence.

FODMAP malabsorption
Although all FODMAPs are poorly absorbed, the anatomical reasoning underlying the incomplete or complete lack of absorption differs among FODMAPs.
Fructose is a hexose sugar being increasingly consumed in its monosaccharide form as an added sweetener and in its more natural forms such as fruit juice 68 .There is no clearly established fructose malabsorption mechanism 68 and most of the understanding of fructose transport has been based on animal studies 69  abdominal pain and diarrhea is it referred to as "lactose intolerance" 71 .
FOS or Fructans are oligo-and polysaccharides of fructose with a glucose terminal end 16,72 . They are classified according to their bonds as inulins (β1-2 bonds) or levans (β2-6 bonds) with most dietary sources coming from inulins 16 . When an inulin has <10 degrees of polymerization (DP) it is referred to as a fructo-oligosaccharide, whereas >10 DP is referred to as an inulin 16 . The β-bonds that hold fructose molecules together are unable to be hydrolyzed by human digestive enzymes, thus theoretically FOS travel unabsorbed in all humans, resulting in more than 90% of fructans reaching the colon 16 . That being said, FOS absorption in the human gastrointestinal tracts has not been assessed, studies come exclusively from rat models 73 .
GOS are nondigestible CHO usually composed of 2-10 molecules of galactose and 1 molecule of glucose 74 . The two most common dietary sources are raffinose, comprised of one fructose, one glucose and one galatose molecule and stachyose, which is a raffinose with one an additional galactose 16 . Humans lack α-galactosidase, the enzyme that hydrolyses the galactosidic linkages of stachyose and raffinose to their simple sugar constituents, resulting in minimal absorption in humans 16 .
Polyols are sugar alcohols that include sorbitol, lycasin, malitol and mannitol, 75,76 . Sorbitol and mannitol are six-carbon polyols isomers that are only partly absorbed via passive diffusion across the small intestine epithelium 76 , with a total of 80% ingested reaching the colon 16 .

Intestinal gas production and the hydrogen breath test
Once CHO are malabsorbed in the small intestine they become substrate for bacteria fermentation, which in turn releases gaseous byproduct into the lumen. More

FODMAP malabsorption in healthy adults
The predominant way that diet alters luminal distention is via intraluminal gas production 5 . Even in healthy individuals, FODMAPs are malabsorbed 12 , shown using breath hydrogen testing to compare the prevalence of CHO malabsorption between functional GI disorder (FGID) patients and healthy subjects. Barrett et al. 17 found that 34% of healthy people (n=71) malabsorb fructose compared with 45% of those with FGID (n=201) when given 35 g of fructose prior to breath hydrogen testing (malabsorption was defined as >10 ppm). In that same study 17

Importance of visceral sensitivity in symptom production
With CHO malabsorption present in both healthy population and FGID patients, and fermentation patterns similar in both populations, a low threshold for visceral pain appears to be the key mediator for gastrointestional symptoms manifestation, particularly abdominal pain 3 . This was demonstrated by Richie et al. 3 who studied the effect of inflating a balloon into the distal colon and compared pain responses between IBS patients (n=67) and healthy adults (n=16). The study found that inflation to 60 mL caused pain in 6% of the control at a mean diameter of 3.8 cm and in 55% of patients with IBS at a mean diameter of 3.4, despite that gut wall tension at that volume appeared to be normal in both groups and gut wall diameter could be further increased. Additionally, in 6% of the controls and 52% of patients with IBS, pain occurred at balloon diameters that could still be increased by 10% or more with further inflation, pointing to a low threshold for visceral pain in patients with IBS compared to healthy adults 3 .

Poorly digestible and Osmotic effect of FODMAP
The participating in the study. Regarding breath hydrogen testing, 50% of the participants did not produce hydrogen. This is understandable given the absence of colonic fermentation in patients without a colon. In the retrospective arm of the study, five of the seven patients had significant improvement in stool frequency (8-4 stools; p=0.02) and consistency as shown by patient self reporting 9 . In addition, patients uniformly reported that reintroduction of prohibited foods worsened symptoms 9 . In the prospective arm of the study, (n=5), no significant change in stool frequency (median 6 to 5 per day; p=ns) occurred. The lack of significance was attributed to acute or chronic pouchitis experienced by three subjects 9 . However the reasoning for the lack of response in patients with inflammation was unclear 9 .
Similarly, a 2010, randomized, single-blinded cross over study where subjects without a colon were given high-FODMAP diets found that effluent liquid output closely related to FODMAP output, clearly demonstrating FODMAPs osmotic properties 6 . This was found by measuring FODMAP output in the stool of subjects. The study consisted of twelve illeostomy patients who for four days consumed diets differing only in FODMAP content. Effluent was collected for 14 hours during the final day of each intervention.

Studies Examining Adherence to low-FODMAP diets
Dietary adherence is crucial to the success of a low-FODMAP diet, however most people do not find the diet easy to incorporate into their life 8,10 . That being said, studies have shown high adherence rates among functional gastrointestinal disorder (FGID) subjects both when all foods are provided in the form of test drinks (>95%) 11 and when asked to follow dietary advice (75.6%) 8 . Adherence among the healthy population who do not experience comparable symptoms has yet to be studied.
Croagh et al. 9 considered change in FODMAP intake, which was used to define adherence in a study examining administration of a low-FODMAP diet. The study was a small, combination retrospective/prospective study, with a total of 15 subjects. In the prospective group, adherence was measured on five subjects using seven-day food records reflecting the intake on the final week of the six week intervention. Adherence was based on total number of "problematic serves" per day, defined as any food that contained >.5 g of free fructose or fructans, >4 g lactose or any sorbitol, which was based on guidelines from a previous study 25 . Each of the five participants reduced the number of problematic serves per day by at least 6.5 serves (P1=8-1.5, P2=12-2, P3=11-0, P4=9-0, P5=12-5) by the end of the intervention. According to Croagh et al. 9 , those with a high baseline intake of dietary FODMAPs and good adherence to the diet responded, while those with a low baseline intake and partial adherence did not 9 .
In another study, de Roest et al. 8 24.4% (n=22) followed the diet up to 3 months, but not anymore; 5.6% (n=5) followed the diet as taught immediately, but less than 50% of the time at the end of the follow-up questionnaire; 4.4% (=4) never followed the diet 8 .

Factors that may contribute to adherence
Gearry et al. 10 conducted a pilot study that explored factors that may contribute to non-adherence to a low-FODMAP diet in patients with Inflammatory Bowel Disease (IBD) based on findings from a previous FODMAP study 11 . Dietary advice consisted of a single one-on-one or group counseling session with a dietitian as well as FODMAP literature and food lists. Adherence was measured using questionnaires via structured telephone interviews regarding FODMAPs consumption as well as specific questions concerning FODMAP-containing foods in order to validate the patient's responses.
According to Gearry et al. 10 70% of IBD patients who suffered from FGID were adherent to advice to follow a low-FODMAP diet. Upon completion of the study, the 72 participants were asked to rate their opinion of the diet on a scale of 0-10 (0=easy, 10=impossible) and obtained median score. Low scores were obtained for the questions 1] how easy was it to implement the diet (median response 3; SD 2.9, range 0-10, interquartile range 0-5), 2] and how easy was it to buy the appropriate foods (median responds 3; SD 2.9, range 0-10, interquartile range 1-4) and 3] how would you rank the overall taste of the diet (median responds 2; SD 2.2, range 0-10 interquartile range 1-4) 10 .
In addition, 44/72 (61%) said that the foods were not available at their usual shops, the higher cost of the diet was thought to be problematic for 46/72 (64%) and the median estimated increase in the cost of food while on the diet was 10% (SD 19, range -10-110%, interquartile range 1-25%).
The de Roest et al. 8

Potential Limitations of a low-FODMAP diet
Several limitations of a low-FODMAP diet pertaining to dietary quality and health benefits have been suggested. Malabsorbed FODMAPs provide multiple benefits including a natural laxative effect due to their osmotic effects 6 , a prebiotic effect with beneficial fermentation by-products 12 and production of a low glycemic response compared to other CHO 21 . Some beneficial by-products of fermentation include short chain fatty acids (SCFA), which may protect against colon cancer as well as promote satiety 89 , and synthesis of B vitamins and vitamin K 90 .

FODMAPs: Low-glycemic index nutrients
FODMAPs are nutrients with lower glycemic indexes 37,38 . Low glycemic index foods are proven beneficial in the treatment and prevention of metabolic syndrome, diabetes and CVD [39][40][41][42] . Although the mechanisms underlying the effects of these foods are not completely understood it is hypothesized that low-GI diets maintain better regulation of blood glucose, which decreases oxidative stress and lowers inflammation 42 .
In addition to immediate response, consumption of low glycemic foods reduces glycemic response at subsequent meals up to 4 hours later 21 .

FODMAPs: Prebiotic Actions
Prebiotics are any nondigestible substances that encourage the growth and activity of favorable intestinal bacteria, known as probiotics, therefore improving the host health 91 and include the FODMAPs FOS, GOS and inulins 92 . Studies have shown that supplementing with FOS, GOS and inulin encourages growth of the beneficial bacteria bifidobacteria, at the expense of less desirable groups of bacteria 93,94 . Beneficial probiotics also include the bacteria lactobacilli; However bifidobacteria are the usual target since these bacteria are more readily altered and more prevalent in the human colon 95 . Bifiobacteria also exhibit a preference for oligosaccharides 95

Dietary quality of Low-FODMAP diets
Only one retrospective study has calculated the diet of free-living subjects who received low-FODMAP dietary advice. Not one study has looked exclusively at healthy adults, changes in intake or looked at overall diet quality as compared to established guidelines. Ostgaard et al. 23  and 34.2±.9 ml, guided; 21.0±6.5 and 16±2.9 ml, unguided; 13.9±5.9 and 14.5 ±4.3 ml respectively) 23 . Fiber however was not assessed as significantly different among the three groups and overall dietary quality was not measured.

CHO malabsorption 's effect on mood
Multiple studies have been conducted linking specific CHO malabsorption to changes in mood or increases in undesirable mood states. Ledochowski  Upon further statistical analysis, individuals with tryptophan concentrations lower than the median (=67.0μM) more often presented with a BDI score above the median (p=.036; Fisher exact test) and when analyses was restricted to fructose malabsorbers, a significant inverse relationship between tryptophan concentration and BDI scores were found both overall (n=35; r=-0.348, p=.043) and when restricted to females (n=24; rs=-0.503, p=.014).

Tryptophan levels and Mood
Although serotonin (5-HT) is often thought of as a neurotransmitter exclusive to the central nervous system (CNS) due to its well-defined role in expression of depression, arousal, pain and other characteristics commonly attributed to CNS functioning, the major source of bioavailability is located in the bowel 102 . Low levels of brain 5-HT can contribute to decreases in mood 103 and are therefore the target of several antidepressants.
The precursor of 5-HT is tryptophan, which is considered an essential amino acid, indicating it cannot be produced internally and must be obtained externally via the diet.
Lowering tryptophan levels through dietary modifications is associated with a postprandial mood-lowering effect 104 .
Ledochowski et al. 98 demonstrated that malabsorption of an individual FODMAP has been associated with decreases in tryptophan levels. According to this study, high intestinal fructose concentrations, as is the case with fructose malabsoprtion, seem to interfere with L-tryptophan metabolism and thus reduce the bioavailability of 5-HT. It was then hypothesized that this could be due in part to a combination of increased transit time and the phenomenon known as the Maillard reaction. The Maillard reaction, which is primarily associated with food science, is a heat-driven process where an amino acid becomes bound to a simple sugar. Ledochowski et al. theorized that malabsorbed fructose results in a fructose-L-tryptophan complex, which is then lost in excretion.
Based on this theory, a diet high in multiple, poorly absorbed CHOs such as a high-FODMAP diet, may lead to reduced levels of the bioavailability of tryptophan and possibly impact mood perception however, proof of concept studies are needed.

Conclusion
Irritable Bowel Syndrome (IBS) is the most commonly diagnosed Gastrointestinal (GI) condition in the United States 48 . In 2012, it was estimated that 10% of Americans meet the diagnosable criteria for IBS, translating to 30 million people 48 .
IBS is an umbrella term that incorporates a spectrum of chronic or recurrent symptoms including abdominal pain, bloating, distension, excessive wind and altered bowel habits when anatomical abnormalities and inflammation have been excluded 12 . A low-FODMAP diet is now considered an effective strategy for managing symptoms of IBS in Australia, with interest expanding across the world 12 . FODMAPs' ability to increase GI symptoms are centered around FODMAPs' three common functional properties; They are 1) poorly absorbed in the proximal small intestine, allowing substrate to reach the distal small intestine and proximal colon 2) small and osmotically-active, which increases the liquidity of luminal content due to osmosis and 3) rapidly fermented by gut microbiota, increasing the amount of gas present in the colon. Dietary adherence is crucial to the success of a low-FODMAP diet, however most people do not find the diet easy to incorporate into their life 8,10 . Several limitations of a low-FODMAP diet pertaining to dietary quality and health benefits have been suggested. Malabsorbed FODMAPs provide multiple benefits including a natural laxative effect due to their osmotic effects, a prebiotic effect with beneficial fermentation by-products 12 and production of a low glycemic response compared to other CHO 21 . Additionally malabsorption of certain FODMAPs has been linked to increases in undesirable mood states [98][99][100] . Not one study has looked exclusively at healthy adults, changes in intake or looked at overall diet quality as compared to established guidelines. A study is needed looking at dietary quality of low-vs. high-FODMAP diets and should consider adherence and other factors that may influence efficacy and potential impact of the diet.

Study Design
The study was done by the Energy Balance Lab (EBL) at The University of Rhode Island (URI) in the spring/summer 2013. It was a randomized, single-blinded, cross-over study comparing two dietary conditions in a free-living setting; a low-FODMAP and a high-FODMAP diet. In order to ensure that the diet was single-blinded, the study was entitled "The Carb Study" and the two conditions were labeled "Diet 1" and "Diet 2" representing the low-FODMAP and high-FODMAP diets respectfully. The diet instruction booklet that corresponded to each dietary condition was developed specifically for this project based on multiple published articles 1,5,23,24 . The selection process was randomized, with a coin flip determining which group the first participant would begin. Each of the two conditions lasted 3 days: Tuesday, Wednesday, and Thursday. An eleven day wash out period where subjects consumed their normal diet separated the two conditions. The subjects had baseline measurements and measurements after completing each diet measured on Tuesday and Friday mornings after a 10 hour fast. In total, there were five visits: an initial assessment (visit 1), two baseline testing (visit 2 & 4), two post-diet testing (visits 3 & 5).

Recruitment
The majority of the subjects were recruited from a list of "Potential Study Volunteers" comprised of adults who were candidates or participants in previous, nutrition-related studies and expressed a desire to be contacted for future studies. A mass email was sent from the EBL team to any adult on this list. In addition to this list, classroom announcements were made in three nutrition classes made up of primarily nutrition students or students in other health-related fields. The estimated attrition rate was expected to be low and was based on a study done by Dr. Melanson (The PI) with a similar demographic and study design (25). In total, 20 participants began the intervention, the attrition rate was 20% and the final sample size was 16. Of the four who did not complete, one subject dropped out due to a hospitalization that involved antibiotic treatment and three subjects did not report to the lab for an appointment.

Subjects
Overall, 18 healthy subjects, free of any gastrointestinal illness completed the study. Gastrointestinal illness included celiac disease, IBS, lactose or gluten intolerance, diverticular disease, colitis such as Crohn's disease or ulcerative colitis or stomach ulcers.
Additional exclusion criteria included any food allergies, being a current smoker, being on a weight loss diet, pregnant or lactating, type 1 or 2 diabetes, adrenal disease, kidney or bladder problems a thyroid disease or currently taking any appetite suppressant medications.

Initial Assessment/Screening
During visit 1, potential participants completed an initial assessment and a screening which assured their status as a healthy adult clear of any GI complications.
Once subjects were declared eligible, demographic measurements and assessment of body fat percentage using the BOD POD Body Composition System (Life Measurement Instruments, Concord, Calif., USA) was performed. BODPODs are used to estimate % fat via air displacement plethysmography (ADP). The procedure has been described in full in a previous study 105

Baseline testing: Postprandial testing
Thirty minutes postprandial, subjects completed the same appetite/discomfort and repeated the same collection methods as fasting measurements described above. A third and final round of testing using the same procedures was conducted 60 minutes postprandial. The break between these three testing points was allocated to subjects receiving dietary instructions for their intervention, and completion of a 24-hour recall.
The total time of these visits was approximately 75 minutes.

Baseline Testing: Diet Instructions and Diet Booklet
During baseline visits, subjects met with a member of the research team educated in the FODMAP diet who provided each subject with a16 page diet instruction booklet labeled either "Diet 1" or "Diet 2". The booklets contained detailed lists of recommended and restricted foods. A brief, 15 minute diet explanation was also provided which included identifying encouraged and discouraged foods, brief tips and reiteration of the impotence of adhering to the diet for the purpose of the study's success. "Diet 1" corresponded to the low-FODMAP diet and "Diet 2" was the high-FODMAP diet.

Post-Diet Testing
An almost exact replica of baseline testing protocol was used for post-diet testing. The only addition was the addition of an "opinion regarding the diet" 6 question mixed VAS and free response questionnaire (Appendix D). The only exemption was that no dietary instructions were provided during the POST-Intervention. At the end of the POST-diet visit, subjects were told to either follow their normal diet (visit 3) or were informed that the study was completed (visit 5). The subjects received a $20 incentive on visit 3 and a $60 incentive on visit 5.

Post-diet testing: Diet opinion
The diet opinion scale (Appendix D) used during the POST-Intervention was developed for this project and had not been used in a previous study. The questionnaire was developed based on a questionnaire created by Gearry et al. 10 using items identified by Shepherd et al. 11 as potential barriers to adhering to a low-FODMAP diet. Of note, since the completion of this study, a questionnaire with similar items has also been used by De Roest et al. 8 .

Additional Questionnaires
In addition to in-lab data collected, subjects were given three days worth of questionnaires assessing appetite, symptoms and mood. The appetite questionnaire used was the same questionnaire described earlier, however the abdominal discomfort scale was omitted. It was filled out pre-and post-meals, mid-afternoon, mid-evening and before bed, for a total of nine times throughout the day. The mood and the symptom questionnaires were filled out daily, before bed. The mood questionnaire used (Appendix D) was a non-validated VAS questionnaire developed by the EBL and used in only one previous study 31 26 . Briefly, pass one included obtaining a quick list of foods consumed in the past 24 hours. In pass two, participants are asked to produce details regarding foods on their quick list including portion sizes and amounts eaten. In pass three the list is recited and participants are asked if any information was forgotten.
Foods that were not in the NDS-R database were cited as "missing foods" and corrected after the interview was completed. Resolution of a missing food usually required finding an NDSR substitute (very similar food or a generic version of a food) in the database and matching that substitute for CHO, protein, fat and kcals. Matching was defined as within 1-3 grams for each macronutrient and within 10 kcals for energy. For some foods, the FODMAP content was considered too variable for a substitution (for example ice cream brands and gluten free products). These foods were emailed to NDSR, who then provided an accurate nutrient breakdown for those items. As mentioned previously, mood and symptom questionnaires were obtained for each day of the 3 day intervention. Participants completed these questionnaires at night, just prior to bed. The scores of interest were the scores obtained on the third day of the intervention. FODMAP intake for that same day was reflected on the 24 hour food recall obtained during the Post-Diet testing. This allowed for comparison of participants' mood scores to 1) their intake of FODMAP, 2) their reported symptoms 3) their HEI-2010 scores on the same day and at the end of the three day intervention. Appendix F show questionnaire data from mod variables that did not make it into the manuscript results.

Statistics
This was a secondary data analysis from a larger study powered on blood glucose.
Statistical analyses were performed using SPSS (v22). All variables met criteria for normality. Baseline comparisons between subjects assigned to the two orders were conducted using t-tests and Χ 2 tests. Primary outcomes (grams FODMAP and HEI-2010 scores) were assessed using separate 2 (treatment) x 2 (time) x 2 (order) mixed factorial ANOVA followed by within-treatment paired t-tests (baseline and intervention Calculation for the HEI-2010 was based off of a protocol developed by the Nutrition Coordinating Center (NCC) at the University of Minnesota based off of methods described in a previous study_ENREF_106 30 . It is important to note that it is only possible to estimate an approximation of the HEI-2010 score using NDSR. Steps: Step one of calculating the total HEI-2010, involved calculating each of the individual index components.
Step two included taking that number and conforming it to the unit of measure used in the index (such as servings converted to cups).
Step three involved comparing intake of each item to the score rubric (table 1).
Step 4 involves summing the individual scores to produce a single HEI-2010 score. Two decimal points were used for every spot. The following items were extracted from output file 9 from NDSR and summed up to give total fruit (servings): Citrus juice, fruit juice excluding citrus juice, citrus fruit, fruit excluding citrus fruit, avocado and similar, fried fruits and fruit-based savory snacks 2.
Total fruit (serving) was than divided by two to produce total fruit (cups) 3.
Total energy expenditure (taking from NDSR output file 04) was first divided by 1000.

4.
Then total fruit (cups) was divided by the results of step 3 5.
The result of step 2 was then multiplied by the optimum total fruit score (5) and divided by the standard for maximum total fruit score (.8) to yield the total fruit score. A maximum of 5 and minimum of 0 was used Whole Fruit

1.
The following items were extracted from output file 9 from NDSR and summed up to give whole fruit (servings): citrus fruit, fruit excluding citrus fruit, avocado and similar, fried fruits and fruit-based savory snacks 2.
Whole fruit (serving) was than divided by 2 to produce total fruit (cups) 3.
Total energy expenditure (taking from NDSR output file 04) was first divided by 1000 4.
Whole fruit (cups) was divided by the results of step 3 5.
The result of step 2 was then multiplied by the optimum whole fruit score (5) and divided by the standard for maximum whole fruit score (.4) to yield the whole fruit score. A maximum of 5 and minimum of 0 was used Total Vegetables

1.
The following items were unconditionally extracted from output file 9 from NDSR and summed up to produce total vegetables (servings): Dark-green vegetables, deep yellow vegetables, tomato, white potatoes, fried potato, other starchy vegetables, other vegetables, friend vegetables and vegetable juice.

2.
Legumes (cooked dried beans) was extracted and added to the total vegetable component score only if the "total protein foods" (including legumes (cooked dried beans)) max score (>2.5 oz eq/1000 kcals) was reached.

3.
Total vegetables (serving) was than divided by 2 to produce total vegetables (cups) 4.
Total energy expenditure (taking from NDSR output file 04) was first divided by 1000.

5.
Then total vegetables (cups) was divided by the results of step 4 6.
The result of step 3 was then multiplied by the optimum total vegetable score (5) and divided by the standard for maximum total vegetable score (1.1) to yield the total vegetable score. A maximum of 5 and minimum of 0 was used Greens and Beans

1.
Dark green vegetables (servings) from output 09 was extracted and used as the greens and beans score 2.
Legumes (cooked dried beans) was extracted and added to the total vegetable component score only if the "total protein foods" (including legumes (cooked dried beans)) max score (>2.5 oz eq/1000 kcals) was reached.

3.
Greens and beans (serving) was than divided by 2 to produce greens and beans (cups) 4.
Total energy expenditure (taking from NDSR output file 04) was first divided by 1000.

5.
Then greens and beans (cups) was divided by the results of step 4 6.
The result of step 3 was then multiplied by the optimum total vegetable score (5) and divided by the standard for maximum total vegetable score (.2) to yield the total vegetable score. A maximum of 5 and minimum of 0 was used Whole Grains

2.
Total energy expenditure (taking from NDSR output file 04) was first divided by 1000.

3.
The result of step 1 was then multiplied by the optimum whole grains score (10) and divided by the standard for maximum total vegetable score (1.5) to yield the whole grains score. A maximum of 10 and minimum of 0 was used.

1.
The following items were extracted unconditionally from output file 9 from NDSR and summed up to give total fruit (servings): milk-whole, milk, reduced fat, milk, low fat and fat free, milk, nondiary, ready-to-drink flavored milk, whole, ready-to-drink flavored milk-reduced fat, ready-to-drink flavored milk-low fat and fat free, sweetened flavored milk beverage power with non-fat dry milk, artificially sweetened flavored milk beverage with non-fat dry milk, cheese-full fat, cheese-reduced fat, cheese-low fat and fat free, cheese-nondairy, yogurt-sweetened whole milk, yogurt-sweetened low fat, yogurtsweetened fat free, yogurt-artificially sweetened whole milk, yogurt-artificially sweetened low fat, yogurt-artificially sweetened fat free, yogurt-nondairy, pudding and other diary deserts, artificially sweetened pudding and other diary deserts, dairy-based sweetened meal replacement/supplement, diary-based artificially sweetened meal replacement/supplement 2.
Frozen diary deserts was also obtained from output file 09, then times by three and added to the score obtained in step one to produce the total diary score 3.
Total energy expenditure (taking from NDSR output file 04) was first divided by 1000.

4.
Then total diary (cups) was divided by the results of step 3 5.
The result of step 4 was then multiplied by the optimum total diary score (10) and divided by the standard for maximum total fruit score (1.3 cups) to yield the diary score. A maximum of 10 and minimum of 0 was used Total Protein Score 1.
The following items were extracted unconditionally from output file 9 from NDSR and summed up to total protein (oz equiv): beef, lean beef, veal, lean veal, lamb, lean lamb, fresh pork, lean fresh pork, cured pork, lean cured pork, game, poultry, lean poultry, fried chicken-commercial entrée and fast food, fish-fresh and smoked, lean fishfresh and smoked, fried fish-commercial entrée and fast food, shellfish, fried shellfishcommercial entrée and fast food, cold cuts, lean cold cuts and sausage, organ meats, baby food meat mixtures, eggs, egg substitute, nuts and seeds, nuts and seed butters and meat alternative.

2.
From the NDSR output file 09, (legumes x2) was added only if the score from step 1 was less than <2.5 oz/1000 kcals.

3.
Total energy expenditure (taking from NDSR output file 04) was first divided by 1000.

4.
Then total protein score (oz) was divided by the results of step 3 5.
The result of step 4 was then multiplied by the optimum total diary score (5) and divided by the standard for maximum total protein score (2.5 oz) to yield the diary score. A maximum of 5 and minimum of 0 was used Seafood and plant protein 1.
The following items were extracted unconditionally from output file 9 from NDSR and summed up to seafood and plant protein (oz equiv): fish-fresh and smoked, lean fish-fresh and smoked, fried fish-commercial entrée and fast food, shellfish, fried shellfish-commercial entrée and fast food, nuts and seeds, nut and seed butters, meat alternative 2.
From the NDSR output file 09, (legumes x2) was added only if the score from step 1 was less than <2.5 oz/1000 kcals.

3.
Total energy expenditure (taking from NDSR output file 04) was first divided by 1000.

4.
Then total protein score (oz) was divided by the results of step 3 5.
The result of step 4 was then multiplied by the optimum total diary score (5) and divided by the standard for maximum total protein score (.8 oz) to yield the diary score. A maximum of 5 and minimum of 0 was used Fatty acids

1.
From NDSR output file 04 the sum of all PUFAs and total MUFAs were extracted and added together 2.
From NDSR output 04 the sum of all SFAs were added together 3.
The result of step 1 was divided by the result of step 2 4.
The following equation was used to determine the fatty acid component score (result of step 3-1.2)*10/1.3. A minimum of 0 and a maximum of 10 was used
Total energy expenditure (taking from NDSR output file 04) was first divided by 1000.

3.
Then total protein score (oz) was divided by the results of step 3 4.

5.
The following formula was then used to yield the final refined grain score ((the result of step 5)*10/2.5). A minimum of 0 and maximum of 10 was used. Sodium 1.
The total amount of sodium (mg) extracted from output 09 from NDSR was obtained and multiplied by 1000 to yield sodium in g 2.
Total energy expenditure (taking from NDSR output file 04) was first divided by 1000. 3.

4.
The following formula was then used to yield the final refined grain ((result of step 3)*10/0.9) a minimum of 0 and a maximum of 10 was used Empty calories 1.
From NDSR output file 04: a.
Total trans fat (g) x 9 kcal/gram= kcal from trans fat c.
From NDSR output file 04, alcohol consumption was determined by the following steps: a.
If alcohol (g) is > than allowable alcohol (g), then: i.
Sum the results of 1-a,-b,-c and 2-b-i 4.
Based on the total energy intake from output file 04, the following equation was used a.
Then, the following equation was used: 100-(result of step 4-a) 6.
The following formula was then used to yield the final empty calorie score: ((the result of step 5-50))*20/31. A minimum of 0 and a minimum of 20 was used Calculating the total HEI-2010 1.
The results of each individual component scores was added together to yield the HEI-2010