A comparative study of cognitive function among children with coeliac disease and healthy controls

Abstract

Coeliac disease (CD) is an immune-mediated systemic disorder triggered by gluten in genetically predisposed patients. The only available treatment is a strict life long gluten-free diet (GFD), which has been linked to a reduced quality of life (QOL) and causes alterations in the gastrointestinal microbiome. Abnormal compositions of the microbiome are now recognized as factors in the pathogenesis of neuropsychological disorders via gut-brain-axis. The aim of this study was to assess the QOL and the mental performance of children and teenagers with CD and compare it to healthy controls (HC). Children between the ages of 6 and 18 years with CD and age-and-sex-matched healthy controls (HC) filled in a questionnaire to assess QOL and performed the Flanker task, a standardized test to assess cognitive performance. A total of 444 children (210 CD patients and 234 HC) were included in the study. CD patients reported feeling statistically significantly more comfortable at school (p = 0.02) and being less exposed to bullying (p = 0.01); otherwise, no difference in QOL and sleep pattern was found compared to HC. The analysis of Flanker task revealed no difference in accuracy (HC: mean 0.97, CI 0.96–0.97; CD patients: mean 0.96, CI 0.96–0.97; p = 0.79), but there was a difference in reaction time (HC: mean 495.4 ms, CI 476.34-514.46; CD patients: mean 514.03, CI 493.68-534.39; p = 0.19). Children with CD in Switzerland have the same QOL as HC. There was a statistically non-significant difference in reaction time, therefore this study suggests that a GFD is not associated with impaired cognitive function.

Background

Coeliac disease (CD) is an immune-mediated systemic disorder triggered by gluten in genetically predisposed patients, with a prevalence of 1–2% in central Europe¹. The only available treatment is a strict lifelong gluten-free diet (GFD).

Older studies showed that sleep disorders are more common in CD patients, which are related to depression, anxiety, and fatigue². Adult CD patients also reported a significant burden of the dietary treatment compared with other chronic medical conditions³, in addition, they complain about social restriction⁴ and stigmatization⁵.

There are nutritional risks associated with a GFD, such as a high intake of fat and cholesterol, a lack of calcium, vitamin D, fiber, vitamin B, and iron^6,7. The absence of gluten in the diet al.so leads to a different composition of gastrointestinal bacteria (microbiota) compared to healthy subjects⁸. In the last decade, a lot of research has been performed on the microbiota and its communication with the brain. Collectively, these pathways are referred to as microbiota-gut-brain-axis (MGBA): The vagus nerve is the most direct route, as different receptors on the vagal afferents detect and transmit stimuli, such as microbial-derived metabolites, their products, and gut hormones (short-chain fatty acids, dietary tryptophan, lipopolysaccharide, polysaccharide A, CCK, 5-HT)⁹. In addition, enteral microbes have been recognized to synthesize neuroactive substances (dopamine, noradrenaline, acetylcholine, histamine, gamma-aminobutyric), which modulate the brain¹⁰. There has been increasing evidence implicating an imbalance of microflora in the development of various neurodevelopmental/psychiatric diseases, such as Alzheimer’s, depression, anxiety, and mental performance¹¹. Since the MGBA communicates in a bidirectional manner, several study groups tried to manipulate the microbiota via supplements with probiotics to combat mental dysfunction. Many studies achieved mitigating shifts in bacteria abundance, but the clinical outcome regarding cognitive performance differed in the various studies¹².

The aim of this study was the comparison of QOL and cognitive function between children and teenagers suffering from CD, who avoid gluten and their healthy peers.

Methods

Between July 2020 and August 2021, all children aged between 6 and 18 years, under the care of the University Children’s hospital of Basel and the Children’s hospital of Aarau, Switzerland, diagnosed with CD according to the Guidelines for diagnosing CD of the European Society for Paediatric Gastroenterology, Hepatology and Nutrition¹³, were informed about the study. Further inclusion criteria were a strict adherence to the diet, defined as a negative transglutaminase for at least one year before participation in the study. Exclusion criteria comprised a diagnosis made less than a year ago, the presence of other chronic diseases, including food allergies, and the intake of regular medication. CD patients and their parents were informed about the study during the regular appointment for a clinical and serological check-up. After obtaining their consent, patients received another appointment to participate in the study.

For every CD-patient an age-and-sex-matched healthy control was recruited from local schools and swimming clubs. CD patients could also bring suitable volunteers Exclusion criteria for healthy controls (HC) were chronic diseases, regular medication intake, food allergies or specific diets.

Firstly, all participants were asked to fill in the questionnaire to examine the QOL. ‘Kidscreen52’¹⁴ was used, a standardized questionnaire containing 52 questions, developed to assess the subjective health and the psychological, mental, and social well-being of children and adolescents covering ten important aspects of life (physical well-being, psychological well-being, moods & emotions, self-perception, autonomy, parent relation & home life, financial resources, peers & social support, school environment, bullying). The answers to all questions were converted into numbers from one to five. In most subgroups, a higher number indicates a more positive answer, except for ‘self-perception’ (questions are like: ‘do you doubt your appearance’) and ‘bullying’, where a lower number indicates also less exposure to bullying. Participants were also asked to fill in the ISI (insomnia severity index), a 7-item self-report standardized questionnaire assessing the nature, severity, and impact of insomnia¹⁵.

Participants completed a computerized Flanker task¹⁶ to assess cognitive performance. In this cognitive task, the target was flanked by non-target stimuli, which correspond either to the same directional response as the target (congruent flankers) or to the opposite response (incongruent flankers). In our child-friendly variant, participants were presented with five fish in a row and were asked to press the correct button according to the orientation of the target (central fish). Participants were instructed to respond as accurately and as quickly as possible. The two different conditions were presented with equal probability. Visual stimuli were shown for 200 ms against a black background, followed by a blank inter-trial period of random length between 900 and 1300 ms. Participants completed one practice block and two test blocks with 40 trials each. Flanker task is a reliable, well-recognized neuropsychological test for assessing executive function¹⁷. Mean reaction times (on response-correct trials) and accuracy rates were calculated for congruent and incongruent trials.

To have consistency, the interviews and instructions for Flanker task were conducted by the same three students. Incomplete patient data sets were excluded.

Statistical analysis

A priori power analysis, based on an alpha level of p <0.05 for an ANCOVA with 3 covariates (age, sex and average daily moderate-to-vigorous physical activity) concluded that a total of 146 participants (n = 73 patients with CD) are required to reach 85% power. Prior to man analyses, Shapiro-Wilk test was employed to verify the Gaussian distribution of the data.

Between-group differences were tested via ANCOVA that controlled for age and sex. The group effect was reported along with partial eta². Additionally, partial correlations (controlling for age and sex) were calculated between QOL outcomes and performance on the Flanker task. For all comparisons, the level of significance was set to p < 0.05.

SPSS 28.0 was used for statistical analyses (IBM Corporation, Armonk, NY, USA).

Results

Over 450 patients with CD meeting the inclusion criteria were invited to participate in the study. A total of 210 CD patients (38.6% male, 61.4% female) and 234 healthy controls (50.9% male, 49.1% female) were included in the study (for further patient’s characteristics see Table 1). With regard to QOL, ANCOVA showed only statistically significant differences in two aspects: Bullying and the school environment. CD patients reported to feel statistically significant more comfortable at school (p = 0.02) and to be less exposed to bullying (p = 0.01) (see Table 2; Fig. 1).

Table 1.

Patient’s characteristics.

	Patients with coeliac disease			Healthy controls
	Females	Males	Total	Females	Males	Total
n	129	81	210	115	119	234
Mean age (years)	12.6	12.7	12.6 (range: 6–19)	13.0	13.1	13.0 (range: 6–20)
Mean weight (kg)	43.7	48.0	45.4 (range: 18.2–120)	47.2	50.6	48.8 (range: 18–95)
Mean length (cm)	151.6	155.9	153.0 (range: 116–180)	156.1	159.7	157.8 (range: 115–193)

Table 2.

Results of quality of life (qol) and flanker task.

	Healthy controls (HC)			Coeliac disease (CD) patients
	Mean	Standard deviation (SD)	Confidence interval (CI)	Mean	Standard deviation (SD)	Confidence interval (CI)	p	partial eta2
Physical well-being	4.0	0.05	3.9–4.1	4.0	0.05	3.9–4.1	0.30	0.003
Psychological well-being	4.0	0.05	3.9–4.1	4.1	0.05	4.0-4.2	0.07	0.008
Moods & emotions	4.3	0.03	4.2–4.4	4.4	0.04	4.3–4.4	0.12	0.006
Self-perception	2.8	0.03	2.7–2.9	2.8	0.03	2.7–2.9	0.36	0.004
Autonomy	4.1	0.05	4.0-4.2	4.2	0.05	4.1–4.3	0.07	0.002
Parent relation & home life	4.5	0.04	4.4–4.6	4.5	0.04	4.4–4.6	0.60	0.008
Financial resources	4.5	0.05	4.4–4.6	4.6	0.05	4.5–4.7	0.13	0.001
Peers & social support	4.1	0.04	4.1–4.2	4.2	0.04	4.2–4.3	0.08	0.006
School environment	4.0	0.05	3.9–4.1	4.2	0.05	4.1–4.3	0.02	0.008
Bullying	1.4	0.04	1.3–1.4	1.3	0.04	1.2–1.4	0.01	0.013
Insommnia sleep index (ISI)	14.1	0.27	13.6–14.6	14.4	0.28	13.9–14.9	0.35	0.016
Accuracy in probability	0.97	0.005	0.96–0.97	0.97	0.005	0.96–0.98	0.79	0.000
Reaction time in ms	495.4	9.7	476.3-514.5	514.0	10.3	493.7-534.4	0.19	0.009

Fig. 1.

Quality of life of children with coeliac disease and healthy controls.

There were no differences in the quality of sleep between patients and HC. The analysis of Flanker Task showed no difference between the two groups in accuracy (HC: mean 0.97, SD 0.005, CI 0.96–0.97; CD patients: mean 0.97, SD 0.005, CI 0.96–0.98; p = 0.79) and only a slight difference in the reaction time (HC: mean 495.4 ms, SD 9.7, CI 476.3-514.5; CD patients: mean 514.0, SD 10.3, CI 493.7-534.4), statistically not significant (p = 0.19).

Partial correlations -corrected for age and sex- revealed a relation of accuracy on incongruent trials with positive emotions (p = 0.027) and sleep (p = 0.027), while accuracy on congruent trials was associated with a positive family environment (p = 0.017).

Discussion

The main findings of this study are-except for the school environment and bullying- no differences between the QOL and sleep of CD patients and HC. Whereas groups only differed minimally in their performance on the Flanker task (minimal longer reaction time for CD patients, statistically not significant), accuracy was correlated with positive emotions, sleep and a positive family environment.

The QOL for CD patients might be the same as for HC, as the consumption of gluten-free foods has significantly increased. From 2011 until 2016, the money spent on such groceries in North America has doubled¹⁸. This is in line with a British follow- up survey, which found, over a three-year period an increase in self-reported gluten sensitivity by over 250% (from 12.9% in 2012 to 32.8% in 2015). Clearly, this does not reflect an increase in CD or gluten sensitivity diagnoses, but it has become a lifestyle. In 2014, a North American market research found that 65% of people consuming gluten-free food believe that this diet is healthier in general¹⁹. A GFD has also been advocated for athletes²⁰ and autistic patients²¹ to increase physical and mental performance. However, this has never been scientifically proven, but the increased demand naturally leads to a greater supply of gluten-free products, which in turn, can enhance the QOL of CD patients.

Regardless of the motivations behind abstaining from specific food items, a discernible trend has emerged over the past decade wherein children progressively omit certain foods from their dietary intake, often due to factors like allergies, intolerances, religious beliefs, or specific diets^22–24. Consequently, this normalization of selective eating habits among children and adolescents might facilitate adherence to prescribed diets, as it establishes a societal understanding that not everyone consumes all food types. The easier access to gluten-free, now palatable products, along with the fact that adhering to a specific diet as a child/teenager is no longer considered exotic or conspicuous, accounts for the absence of distinction between individuals with CD and HC. It is conceivable that GFD itself enhances QOL, particularly given that symptoms have improved and there are no adverse medical side effects associated with the diet.

We cannot correlate the fact that children with CD even experience less bullying and perceive school more positively than HC with the condition itself. It is well known, that children whose parents are involved tend to have a better QOL and perform better at school²⁵. It is possible that highly committed parents visit the doctor more often/quickly, resulting in a higher detection rate of CD in this group, which might explain this finding. In this study cohort, the quality of sleep was the same for patients with CD and HC, but children/adolescents with a regular medication have been excluded from this study, which makes it naturally difficult to properly assess this aspect in this cohort.

The Flanker task, over its existence, has undergone adaptations and refinements aimed at the objective measurement of mental performance²⁶. Its role in research and theoretical advancements pertaining to performance dynamics within and across trials is substantial. Notably, the U.S. National Institutes of Health have formally endorsed it as a standardized assessment tool to gauge attention and control abilities across various life stages²⁷.

By evaluating reaction time and accuracy, this task offers an objective means to quantify both the ability to focus (distinguish between conflicting stimuli) and the broader construct of executive function. The cognitive capacity known as executive function is a cognitive ability managing complex or conflicting information to achieve specific goals, and it holds significant importance in development. This skill correlates with enhanced academic performance, improved social aptitude, reduced instances of aggression and disruptive conduct, and ultimately fosters more favorable outcome during adulthood²⁸.

In our cohort we found no significant differences in performance on the Flanker task between CD patients and HC. Existing literature on this is very sparse.

Benefits of human-microbe symbiosis are known to extend to mental executive function via bidirectional communication between gastrointestinal microbes⁹. Recent studies indicate that individuals predisposed to CD already have a different microbiome (more Bacteroides, Prevotella and less anti-inflammatory bacteria such as Bifidobacteria and Lactobacilli²⁹, which might be involved in the development of the disease. In addition, the exclusion of gluten from the diet al.so leads to an alteration in the composition of gastrointestinal microbiota; particularly, the relative abundance of Bifidobacteria diminishes consistently after a low-gluten diet³⁰. These marked alterations in the microbiome are likely to elicit an interaction with the brain. A big American cohort study of over 13’000 women found no evidence of any association of long- or short-term gluten intake with cognitive function³¹. It is of importance to notice, that first of all, this was an online-survey/test and secondly, from a dietary perspective, substantial variations exist amongst gluten-free foods: In addition to naturally gluten-free products such as fruits, vegetables, and dairy, there are also industrially produced gluten-free products, which have been proven to be deficient in fibers, proteins, folate, iron, and zinc, while higher in fat and sugars³². Wheat and barley are a main source for oligofructose, which is a critical fructan prebiotic. A reduced intake of it has been shown to create a shift in the diversity of the microbiome (lower bifidobacterium and lactobacillus, increasing enterobacteriaceae), which leads to diminishing short chain fatty and organic acids and ultimately induces a significant lower production of cyto- and chemokines³³ and severely affects the gut-brain-axis⁹.

Numerous studies consistently document a correlation between CD and an array of behavioral and psychiatric disorders, encompassing anxiety, ADHS, mood disorders, sleep disturbances, and notably, depression, which exhibits the strongest association^34–36. However, this correlation was not evident within our paediatric cohort, where individuals with CD and HC reported comparable levels of QOL and sleep patterns. It is possible that a measurable decline in cerebral execution, manifested as delayed reaction time in the Flanker task in children with CD, a result of an altered microbiome, may serve as an early indicator of gradual changes in mental acuity. On the other hand, one could argue that a gluten-free diet improves reaction time, which might have been slower, but has not been determined in patients with untreated CD. However, this remains a hypothesis, particularly as the participants’ microbiota were not examined. It is also important to bear in mind that the microbiota is the result of various factors, and that in this study we only examined a very small, controllable factor that could influence a possible change in the microbiota by distinguishing between GFD and gluten-containing diets. Croall et al.³⁷ examined QOL and cognitive function in adult patients with an established and with a new diagnosis of CD and compared them to HC: They found a statistically significant worse performance of verbal and visual memory in both, new and established CD patients compared to HC. Further, worse QOL outcomes were observed, whilst newly diagnosed patients reported lower vitality and established patients had worse bodily pain. Sub-analyses were made and stratifying by disease duration, the pattern of findings suggests, that cognitive deficit is established at the point of diagnosis, and though it remains generally stable after this point, some further decline may be expected. It is unclear, whether this is a phenomenon only observed in adult patients, as no similar pediatric study has been published and we did not perform sub-analysis regarding the duration of CD. A study group from the UK³⁸ matched adult biobank participants with CD with HC and compared cognitive tests, mental health screenings and MR imaging. They found that CD patients had significant deficits in reaction time, indications of anxiety and health-related unhappiness und white matter changes in the brain. Therefore, authors concluded that CD is associated with neurologic and psychological features. This is also a study of adult patients. A similar study involving children is not available. The duration of the illness may be a factor in the manifestation of psychomotor symptoms.

Our study’s primary strength lies in the objective assessment of mental performance, elucidating the clinical ramifications of the modified gut microbiome in children adhering strictly to a gluten-free regimen due to CD, which has not been done before. Although the participant cohort was extensive, the study is limited in that precise dietary history details were not documented. In retrospect, we also regret not recruiting a third cohort of patients with CD, who did not follow a gluten-free diet (yet), as this would have enabled us to answer the clinical question in an even more scientific manner.

An association between emotions, sleep, family environment- as found in our study- is not surprising, as it is well known that good social support buffers stress and increases mental performance³⁹.

Conclusion

Children in Switzerland suffering from CD have the same QOL as HC.

Our findings suggest that patients with CD on a strict GFD have the same cognitive function as HC. This represents the first study of its kind; further studies should objectively assess the quantity and specific types of micro- and macronutrients consumed by patients. This will facilitate a more definitive conclusion regarding the correlation between a GFD and mental performance.

Abbreviations

CD

Coeliac disease

GFD

Gluten-free diet.

HC

Healthy controls.

MGBA

Microbiota-gut-brain-axis.

QOL

Quality of life.

SD

Standard deviation

Author contributions

CL and RF formed the concept of the study. LP, LS, BT and HKcollected all data. SL performed statistical analysis. CL wrote the first draft of the manuscript. All authors reviewed the manuscript.

Funding

no funding was received.

Data availability

The dataset used is available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Consent for publication

- Not applicable.

Ethics approval and informed consent

The present study was approved by the local ethical committee (Ethics committee of Northwest Switzerland, EKNZ, trial number 2019 − 01095). Furthermore, the study was conducted in accordance to the ethical principles laid down in the Declaration of Helsinki and its later amendments. Informed consent has been obtained by all participants over the age of 16 years and by parents or legal guardians for participants under the age of 16 years.