What Is Known/What Is New
What Is Known
- Children with celiac disease may be at risk of nutritional compromise because of the restrictiveness and expense of the lifelong gluten-free diet they follow.
- Previous studies evaluating the dietary intake of children with celiac disease have reported conflicting results but limited by comparisons with non- or inadequately matched controls.
What Is New
- In the first study using healthy controls matched by age, sex, and socioeconomic status, children with celiac disease consume more calories and dietary fat than controls.
- Children with celiac disease have lower weight and body mass index z-scores than nonceliac disease controls.
Celiac disease (CD) is a gluten-induced, immune-mediated enteropathy (1,2). Treatment is a lifelong gluten-free diet (3), which poses challenges. Gluten-free products may have different nutritional compositions to their gluten-containing equivalents, such as higher fat (4,5) and less fiber content (5), often because of the ingredients replacing the roles of gluten-containing cereals in food (6,7). Furthermore, the range of available gluten-free products is often limited and expensive (8,9). Therefore, achieving recommended intakes for macro- and micronutrients may be difficult when following a gluten-free diet. Pediatric studies in the past 2 decades on the nutritional adequacy of the gluten-free diet have differed in their findings (10–21). These studies have varying designs and are limited by small sample sizes or lack of appropriately matched controls. Only 1 study compared cases with controls matched by multiple factors (age, sex, and body mass index [BMI]); however, this study focused on older children and young adults ages 10–23 years in Spain (20). To our knowledge, no study has investigated whether socioeconomic factors affect dietary intake in CD despite the expense of gluten-free products (8,9).
Evidently, children with CD may have compromised nutritional intake, which could negatively impact their growth and development. Therefore, this study aims to evaluate the nutritional intake of children with CD following a gluten-free diet, including energy, macronutrients, sugar, fiber, and micronutrients, compared with healthy controls (HC). As a secondary aim, the relationship between dietary composition and socioeconomic factors was also investigated.
MATERIALS AND METHODS
This study was designed as a cross-sectional, case-controlled study of children with established CD and HC matched by age, sex, and socioeconomic status at Sydney Children's Hospital, Randwick, Australia. Sydney Children's Hospital is a tertiary referral center for the state of New South Wales (the most populous state in Australia and over 3-fold greater in size than the United Kingdom) and receives children from both rural and city locations throughout the state. This research was approved by the Sydney Children's Hospitals Network Human Research Ethics Committee (LNR/14/SCHN/561) as part of the Dietary Intake Study in Children (DISH) on January 14, 2015.
Eligible celiac subjects were identified at the time of review at scheduled gastroenterology clinics, or via the hospital laboratory database. All CD subjects who met inclusion criteria were consecutively approached for recruitment. They were either approached by a research team member in person or were contacted via phone after a formal letter of invitation was mailed out. Subjects were recruited between January 2015 and August 2017. Inclusion criteria were: children ages 2 to 18 years at recruitment with histologically confirmed diagnosis of CD, and adherent to a gluten-free diet for at least 6 months. Children were excluded if they: were receiving enteral or parenteral nutrition, had any comorbid conditions apart from CD, which required further dietary restrictions or if they were voluntarily following any specific dietary pattern, for example, a vegetarian diet, or if they were found to have an estimated energy intake less than 50% or greater than 200% of reference values for age and sex (22). Healthy children matched for age (+/−1 year), sex, and socioeconomic category (assessed through the Australian Bureau of Statistics’ Socio-Economic Indexes for Areas [SEIFA] codes) were recruited as controls through hospital-based sources, previously reported in a separate study (23), and community-based sources. Hospital-based sources included children related to or known to hospital staff and children attending fracture clinic for playground injuries and fractures. Children were also sourced from the community with snowballing techniques. Celiac subjects referred acquaintances who were then recruited with formal letters of invitation and subsequent contact over phone. Healthy controls were defined as children without any acute or chronic diseases, which would alter their dietary pattern. Healthy controls were excluded if they were receiving enteral or parenteral nutrition, had any comorbid conditions requiring dietary restrictions or if they were voluntarily following any specific dietary pattern, or if they were found to have an estimated energy intake less than 50% or greater than 200% of reference values for age and sex (22).
After receiving written informed consent, contact details were obtained for distribution of electronic socioeconomic and dietary surveys. Consent to follow-up was obtained from each subject in the event that the surveys were not returned. On this occasion, at least three separate attempts were made to contact the subject before they were classified as having dropped out of the study.
Dietary intake was assessed through the Australian Child and Adolescent Eating Survey (ACAES), a 120-item semi-quantitative food frequency questionnaire (FFQ) validated for use in Australian children ages 2 to 18 years (24). The ACAES evaluates the frequency of consumption of different foods in the preceding 6 months. Participants were provided with standardized instructions before completing the survey online. Nutrient intake data was extracted using FoodWorks (Version 3.02.581) and the following databases: Australian AusNut 1999 database (All Foods) Revision 14, and AusFoods (Brands) Revision 5 (Xyris Software (Australia) Pty Ltd, FoodWorks Professional Version 3.02.581. 2004: Brisbane, Australia).
Weight, height, and BMI were obtained for all subjects at the time of recruitment using standardized measures. These values were compared with reference standards from the National Centre for Health Statistics to calculate weight-for-age, height-for-age and BMI-for-age z-scores (25).
Socioeconomic status was quantified by calculating the Index of Relative Socioeconomic Disadvantage (IRSD) using each participant's residential postcode and the Australian Bureau of Statistic's SEIFA database (26). The IRSD is a socioeconomic index that combines indicators of relative social and cultural disadvantage within an area, including level of education, type of employment and percentage of unemployment, household income, prevalence of single parent households, prevalence of poor English skills amongst households, the presence of overcrowding, and the prevalence of disability. A score ranging from 1 to 10 is then generated. A low score is indicative of relative socioeconomic disadvantage, whereas a high score denotes relative socioeconomic advantage. In this study, relative socioeconomic disadvantage was defined as IRSD 1 to 5, and relative socioeconomic advantage was defined as IRSD 6-10.
Statistical analysis of all outcome measures was performed cross-sectionally using SPSS Statistics (Version 24, IBM Corporation, Armonk, NY). R (Version 3.4.2) was used to generate graphs. Continuous variables were analyzed using an independent t-test or Mann-Whitney U test for normal and nonnormal distributions, respectively. Correlation between continuous variables was assessed using a Pearson or Spearman correlation for normally and nonnormal distributions, respectively. Categorical data was analyzed using a chi-squared test. When accounting for multiple comparison bias, a false discovery rate (FDR) correction using the Benjamini-Hochberg method was applied to the P-value and is presented as a q-value. For all analyses, a P-value or q-value (when applicable) of <0.05 was considered statistically significant.
The total number of eligible subjects recruited to this study was 85 CD subjects and 85 matched HC. Twenty children (along with their matched pair) were subsequently excluded because of the following: absent or incomplete socioeconomic or dietary data (n = 15), or implausible energy intakes (n = 5). A final number of 65 CD subjects and 65 HC were available for analyses.
The mean (standard deviation [SD]) age of the participants was 10.2 (3.6) years for celiac cases and 10.1 (3.7) years for HC (P = 0.96), and in each group, 38.5% were boys. For celiac cases, the mean (SD) duration on a gluten-free diet for was 3.6 (2.8) years (range: 0.5–12.6). A summary of the participant characteristics is provided in Table S1 (Supplemental Digital Content 1, http://links.lww.com/MPG/B821). Seven of the 65 participants had been on a gluten-free diet for greater than 6 months but less than 1 year.
Energy, Macronutrient, Sugar, and Fiber Intake
Children with CD had a higher energy intake than HC (2413.2 (489.9) and 2190.8 (593.5) kcal/day respectively, P = 0.02) (Table 1). For macronutrients, children with CD had higher intakes of fat compared with HC (818.1 ± 180.9 and 714.3 ± 212.2 kcal/day, respectively, q = 0.018) (Table 1). Intakes of subtypes of fat (saturated, polyunsaturated, and monounsaturated) were also higher in children with CD than in HC. For the other macronutrients (i.e. protein and carbohydrates), there were no significant differences between children with CD and HC (Table 1). The relative contribution of protein, fat, and carbohydrate to total energy intake and the relative percent contributions of saturated, polyunsaturated, and monounsaturated fat were similar with no significant differences observed. There were also no differences in sugar and fiber intake between children with CD and HC (Table 1).
There were no differences in micronutrient intake between children with CD and HC following FDR correction (Table 2). In both groups, intake of micronutrients met Recommended Dietary Intake (RDI) or average intake (AI) in most children (Table 2).
The distribution of IRSD amongst the celiac and control groups are demonstrated in Figures S1 (a) and (b), respectively (Supplemental Digital Content 2, http://links.lww.com/MPG/B821). There were no differences noted in the various dietary measures between children with celiac disease of relative socioeconomic disadvantage (IRSD 1–5) and relative socioeconomic advantage (IRSD 6–10) (see Table S2 [Supplemental Digital Content 3, http://links.lww.com/MPG/B821] for energy, macronutrients, sugar and fiber, and Table S3 [Supplemental Digital Content 4, http://links.lww.com/MPG/B821] for micronutrients). There were also no differences noted in dietary intake of HC subjects of relative socioeconomic disadvantage (IRSD 1–5) and relative socioeconomic advantage (IRSD 6–10) (see Table S4 [Supplemental Digital Content 5] for energy, macronutrients, sugar and fiber, and Table S5 [Supplemental Digital Content 6] for micronutrients).
Children with CD had significantly lower mean (SD) weight z-scores from HC (−0.06 [1.05] and 0.47 [0.96], respectively, P = 0.003) as well as significantly lower mean (SD) BMI z-scores than HC (−0.02 [0.88] and 0.41 [1.09], respectively, P = 0.02). There was no significant difference in height z-scores between CD and HC subjects (0.01 [1.28] and 0.41 [1.17], P = 0.07) (Table S1, Supplemental Digital Content 1, http://links.lww.com/MPG/B821). For children with CD, there was no correlation between weight z-scores and the duration a patient had been on a gluten-free diet (Spearman correlation coefficient = −0.05, P = 0.7) (Fig. 1). When comparing children with CD who had been on short- and long-term gluten-free diets (less than 3.5 vs ≥3.5 years of gluten-free diet), there were no differences in weight and weight z-scores, height and height z-scores or BMI z-scores (see Table S6, Supplemental Digital Content 7, http://links.lww.com/MPG/B821).
To our knowledge, this is the first study to evaluate the dietary intake of children with CD and HC matched for age, sex, and socioeconomic factors, and to explore associations between dietary intake and socioeconomic factors in this population. The main findings were that children with CD had lower weight and BMI z-scores compared with HC despite having higher energy and fat intake. The intake of other macronutrients and micronutrients was comparable between CD children and HC, and socioeconomic factors did not affect the dietary intake of children with CD. Our study raises potential concern as it may indicate that changes in body composition and growth in untreated CD patients caused by malabsorption and undernutrition do not completely reverse after establishment of a gluten-free diet.
Our findings agree with a number of previous studies reporting higher energy intake (15,18) and higher fat intake (13,16,19) in children with CD compared with HC. One separate study found no difference in energy intake between children with CD and HC but matching of HC was only performed for age (13). Four studies from the past 2 decades found fat intake comparable between children with CD and HC but similarly did not use age- and sex-matching (10,11,17,18). The lack of differences in the relative fat or fat subtype intakes (saturated, polyunsaturated, and monounsaturated fats) between children with CD and HC suggests that the distribution of energy intake across the fat subtypes was similar between the 2 groups (ie, children with CD likely ate similar fat-containing food items, but they ate more). This is in keeping with the findings of previous studies (15,18,20).
A number of previous studies have reported no difference in intakes of protein (11,17,18,20) and carbohydrates (17,18,20) between children with CD and HC, which was concordant with our findings. This is not, however, consistent within the entirety of the literature. Four studies showed that CD children consumed less protein than HC (10,12,16,19), although none matched cases and controls by age or sex. Two studies found that carbohydrate intake was higher in CD children than HC (12,15). One study, however, used data from other dietary studies as a “control” rather than recruiting their own cohort of healthy controls (12), whilst the other had a small sample size (18 CD and 18 HC) (15).
Only three studies have examined sugar intake amongst children with CD and HC (17,18,20). Two of these studies (17,18) found no difference, which agreed with our study findings. The other study by Babio et al (20) in 2017 evaluated dietary intake in a study population with a mean age of 15.3, which may represent an older demographic whose dietary choices are more independent. By comparison, our study population had a mean age of 10.1 years, and therefore, we likely had a higher proportion of subjects whose dietary choices were primarily influenced by their parents. This may explain the discrepancies between our findings. In this study, the lack of difference in fiber intake agreed with previous studies (10,15–17,20), even though historically, children on gluten-free diets were thought to be at greater risk of insufficient fiber intake because of the low fiber content of refined flours in gluten-free products (27).
We evaluated micronutrient intake firstly by comparing intake between children with CD and controls and secondly by examining how well children in both groups satisfied micronutrient requirements (EAR or AI) for their age. This study found no difference in micronutrient intake between children with CD and HC, and reassuringly, intake of all micronutrients largely satisfied EAR or AI in both groups. Discrepancies in methodology make it difficult to compare our findings to previously published dietary studies. Some previous studies have only compared the median or mean intake of a micronutrient to the recommended intake for that country to determine whether children with CD and HC were meeting requirements (12,15). Two studies only compared micronutrient intake between children with CD and HC without comparing intake recommendations, and therefore, lacked information about dietary quality (16,18).
Our study suggests that socioeconomic factors do not affect dietary intake in children with CD. The elevated consumer demand for gluten-free foods in the general population may have increased the availability and reduced the cost of gluten-free products. As study recruitment was based in Australia and dietary patterns were based on contemporary food sources available in Australia, these results may not be generalizable to other countries.
We observed that children with CD had lower weight and BMI z-scores compared with HC, despite consuming more calories, through having a higher fat intake. These findings are consistent with an Italian study, which specifically examined weight, height, and BMI in 71 adults with CD and age- and sex-matched controls (28). Despite being in clinical, biochemical, and histological remission and consuming a strict gluten-free diet for at least 2 years, celiac cases had lower weight and BMI compared with controls (28). This suggests that contrary to previous literature (29,30), changes in body composition and growth that occur in untreated CD patients from malabsorption and undernutrition may not completely reverse despite disease remission. Further investigation is warranted into other factors that may influence nutritional outcomes. A potential contributor is the gut microbiome, which can strongly influence host nutrition and metabolic capacities (31). Children with untreated CD have been shown to have imbalances in duodenal and fecal microbiota compared with healthy controls that are only partially restored after a long-term gluten-free diet (32).
This study was limited by being single-center in nature. Subjects recruited from the gastroenterology clinics in this study, however, constituted a diverse group of children from across New South Wales, Australia's most populous state. All dietary assessment tools have limitations and strengths. Food frequency questionnaires (FFQ), such as the one used in this study, may overestimate intake, and are associated with difficulties estimating portion sizes and recall bias (33–35). We also acknowledge that the FFQ used in this study was constructed to assess normal dietary habits, and coding for gluten-free products and the ability to distinguish between sources of fiber intake are important considerations for future surveys. Their low respondent burden, however, makes them an effective data collection tool and they can summarize usual dietary intake over extended time periods (33,35). Furthermore, the ACAES FFQ utilized in this study was validated for Australian children (24). Finally, weight and height measurements where possible were measured by a doctor, nurse, or member of the research team. A proportion of the children, particularly the healthy controls, however, participated in self-measurement. These participants were provided standardized instruction sheets describing proper measurement techniques.
To conclude, this study has found that children with CD consume more calories through consuming more dietary fat than their nonceliac peers, while maintaining similar intakes of other macronutrients and micronutrients. Reassuringly, despite the restrictions and expense of the gluten-free diet, dietary intake does not appear to be impacted by socioeconomic factors. Despite consuming more calories and fat than their matched HC, children with CD, however, had lower weight and BMI z-scores.
We are grateful to all the participants and their parents for their willingness to contribute to this study. We thank Erin Brighten (South Eastern Area Laboratory Services) for her help in participant identification.
1. Niewinski MM. Advances in celiac disease and gluten-free diet. J Am Diet Assoc
2. Quero JS, Jaime BE, Martínez AR, et al. Nutritional assessment of gluten-free diet. Is gluten-free diet deficient in some nutrient? Anales de Pediatría (English Edition)
3. Hill ID, Bhatnagar S, Cameron DJ, et al. Celiac disease: working group report of the First World Congress of Pediatric Gastroenterology, Hepatology, and Nutrition. J Pediatr Gastroenterol Nutr
2002; 35: (Suppl 2): S78–S88.
4. Kulai T, Rashid M. Assessment of nutritional adequacy of packaged gluten-free food products. Can J Diet Pract Res
5. Miranda J, Lasa A, Bustamante M, et al. Nutritional differences between a gluten-free diet and a diet containing equivalent products with gluten. Plant Foods Hum Nutr
6. Houben A, Höchstötter A, Becker T. Possibilities to increase the quality in gluten-free bread production: an overview. Eur Food Res Technol
7. Wu JH, Neal B, Trevena H, et al. Are gluten-free foods healthier than non-gluten-free foods? An evaluation of supermarket products in Australia. Br J Nutr
8. Singh J, Whelan K. Limited availability and higher cost of gluten-free foods. J Hum Nutr Diet
9. Stevens L, Rashid M. Gluten-free and regular foods: a cost comparison. Can J Diet Pract Res
10. Mariani P, Viti MG, Montuori M, et al. The gluten-free diet: a nutritional risk factor for adolescents with celiac disease? J Pediatr Gastroenterol Nutr
11. Rujner J, Socha J, Syczewska M, et al. Magnesium status in children and adolescents with coeliac disease without malabsorption symptoms. Clin Nutr
12. Hopman EG, le Cessie S, von Blomberg BM, et al. Nutritional management of the gluten-free diet in young people with celiac disease in The Netherlands. J Pediatr Gastroenterol Nutr
13. Ferrara P, Cicala M, Tiberi E, et al. High fat consumption in children with celiac disease. Acta Gastro-enterol Belgica
14. Ohlund K, Olsson C, Hernell O, et al. Dietary shortcomings in children on a gluten-free diet. J Hum Nutr Diet
15. Zuccotti G, Fabiano V, Dilillo D, et al. Intakes of nutrients in Italian children with celiac disease and the role of commercially available gluten-free products. J Hum Nutr Diet
16. Kautto E, Ivarsson A, Norstrom F, et al. Nutrient intake in adolescent girls and boys diagnosed with coeliac disease at an early age is mostly comparable to their non-coeliac contemporaries. J Hum Nutr Diet
17. Tsiountsioura M, Wong JE, Upton J, et al. Detailed assessment of nutritional status and eating patterns in children with gastrointestinal diseases attending an outpatients clinic and contemporary healthy controls. Eur J Clin Nutr
18. Alzaben AS, Turner J, Shirton L, et al. Assessing nutritional quality and adherence to the gluten-free diet in children and adolescents with celiac disease. Can J Diet Pract Res
19. Balamtekin N, Aksoy C, Baysoy G, et al. Is compliance with gluten-free diet sufficient? Diet composition of celiac patients. Turk J Pediatr
20. Babio N, Alcázar M, Castillejo G, et al. Patients with celiac disease reported higher consumption of added sugar and total fat than healthy individuals. J Pediatr Gastroenterol Nutr
21. Sue A, Dehlsen K, Ooi CY. Paediatric patients with coeliac disease on a gluten-free diet: nutritional adequacy and macro- and micronutrient imbalances. Curr Gastroenterol Rep
22. National Health and Medical Research Council. Nutrient reference values for australia and new zealand including recommended dietary intakes, Canberra; 2006.
23. Sutherland R, Katz T, Liu V, et al. Dietary intake of energy-dense, nutrient-poor and nutrient-dense food sources in children with cystic fibrosis. J Cyst Fibrosis
24. Watson JF, Collins CE, Sibbritt DW, et al. Reproducibility and comparative validity of a food frequency questionnaire for Australian children and adolescents. Int J Behav Nutr Phys Act
25. Centre for Disease Control and Prevention. Clinical Growth Charts 2017. https://www.cdc.gov/growthcharts/clinical_charts.htm
. Accessed January 2, 2015.
26. Australian Bureau of Statistics. Socio-Economic Indexes for Areas (SEIFA) 2011. 2013.
27. Thompson T, Dennis M, Higgins LA, et al. Gluten-free diet survey: are Americans with coeliac disease consuming recommended amounts of fibre, iron, calcium and grain foods? J Hum Nutr Diet
28. Bardella MT, Fredella C, Prampolini L, et al. Body composition and dietary intakes in adult celiac disease patients consuming a strict gluten-free diet. Am J Clin Nutr
29. Barera G, Mora S, Brambilla P, et al. Body composition in children with celiac disease and the effects of a gluten-free diet: a prospective case-control study. Am J Clin Nutr
30. Rea F, Polito C, Marotta A, et al. Restoration of body composition in celiac children after one year of gluten-free diet. J Pediatr Gastroenterol Nutr
31. Turnbaugh PJ, Ley RE, Mahowald MA, et al. An obesity-associated gut microbiome with increased capacity for energy harvest. Nature
32. Collado MC, Donat E, Ribes-Koninckx C, et al. Specific duodenal and faecal bacterial groups associated with paediatric coeliac disease. J Clin Pathol
33. McPherson RS, Hoelscher DM, Alexander M, et al. Dietary assessment methods among school-aged children: validity and reliability. Prevent Med
34. Burrows T, Truby H, Morgan P, et al. A comparison and validation of child versus parent reporting of children's energy intake using food frequency questionnaires versus food records: who's an accurate reporter? Clin Nutr
35. Magarey A, Watson J, Golley RK, et al. Assessing dietary intake in children and adolescents: considerations and recommendations for obesity research. Int J Pediatr Obes