Benefits of a Gluten-free Diet for Asymptomatic Patients With Serologic Markers of Celiac Disease

Kalle Kurppa; Aku Paavola; Pekka Collin; Harri Sievänen; Kaija Laurila; Heini Huhtala; Päivi Saavalainen; Markku Mäki; Katri Kaukinen


Gastroenterology. 2014;147(3):610-617. 

In This Article


The results of this randomized study showed that screen-detected and even apparently asymptomatic EmA-positive patients benefit from a GFD as measured by extensive clinical, serologic, and histologic parameters. There was some previous evidence indicating improvement in the well being of screen-detected celiac disease patients on a GFD,[10,11,12,13,14] and we showed this now by a vigorous randomized approach. Here, the greatest differences between the groups on dietary intervention could be seen in serology and in mucosal morphology. For ethical reasons no subjects with evident symptoms were randomized and, in fact, the mean baseline quality of life was even better than previously observed in healthy controls.[10] Likewise, the mean laboratory parameters and BMD were within normal range; one cannot expect major improvements in individuals at such an early stage in celiac disease. Nonetheless, in favor of a GFD, there were also significant differences in the GSRS and quality-of-life scores. This randomized approach showed that subjects who thought they were asymptomatic experienced improvement in several objective disease scores upon adopting a GFD. These findings, as exemplified in Table 4 by a marked decrease in most of the GSRS scores on GFD, suggest that the patients may in fact have accepted mild symptoms as normal and recognized them as abnormal only later when on the diet. It is likely that the changes would have been more evident in the symptomatic subjects who were excluded here but commonly are found by active screening.[11,14] In clinical settings many screen-detected patients also may suffer from atypical symptoms not being recognized as celiac disease.[13]

Until now, active screening of celiac disease has been a subject of controversy. Because of the high prevalence and ambiguous clinical picture and existence of effective treatment, screening of at-risk groups or even the whole population has been suggested.[7,8,13,25] Nevertheless, treatment benefit in apparently asymptomatic patients has remained unsolved. Notwithstanding the known positive effects of a GFD, it is expensive and socially isolating and may be detrimental to quality of life. Furthermore, availability of the products is limited and they may have lower nutritional value and palatability than their gluten-containing counterparts. Here, indeed, the SF-36 social functioning score decreased in the GFD group. Then again, besides objective serologic and histologic parameters, self-perceived GSRS scores and anxiety also improved and the majority of subjects showed excellent dietary adherence and were willing to continue on the diet, indicating that it was not considered harmful. There also has been concern that obese celiac patients might gain more weight on a GFD;[26] however, the present and other recent results do not support this concept.[27,28] Furthermore, although not as apparent here, there is evidence that screen-detected patients may have a decreased quality of life and BMD that improves on a GFD.[10,11,12,13] Nevertheless, besides the decrease in social functioning seen here, in some screen-detected patients the celiac disease diagnosis may lead to increased anxiety and health concerns.[10] Given that unrecognized celiac disease may not necessarily increase the risk of malignancy or mortality,[29,30,31] the consequences of possible screening must be weighed carefully in advance on an individual basis. Furthermore, because the present study was conducted in individuals belonging to an at-risk group, the question of population-based mass screening remains a subject for future studies.

As a result of increasing celiac disease screening, a substantial number of seropositive individuals manifest with only mild enteropathy (ie, Marsh grades I–II).[15] Thus far, strict criteria with the demonstration of villous atrophy (Marsh III) have been necessary to ensure a reliable diagnosis of this lifelong and restrictive disorder. However, the modern biomarkers, particularly EmA, have shown almost 100% specificity for celiac disease.[1] Furthermore, we previously showed that most false-positive cases with initially morphologically normal villi (Marsh I–II) eventually will develop mucosal atrophy if they continue on gluten.[32] Moreover, a randomized trial showed that these EmA-positive subjects already benefit from a GFD before villous atrophy (the end stage of the histologic damage) develops.[15] Accordingly, the latest diagnostic criteria of celiac disease have begun to accept milder forms of enteropathy as diagnostic in seropositive patients.[17] These findings suggest that the hitherto used grouped classifications are no longer optimal in the diagnosis of celiac disease. Nevertheless, because serology and symptoms have been shown to be unreliable markers for the mucosal improvement and dietary adherence in celiac disease,[6] we considered it necessary to use histology as the primary outcome in this study; this is likely even more important in screen-detected subjects with mild clinical presentation. To detect even minor changes in morphology, validated and quantitative Vh/CrD was used in this study instead of grouped classification.[33]

In this study, all EmA-positive subjects had celiac disease–associated HLA and all but 1 subject had an increased δγ+ IEL count, which is further proof of the presence of celiac disease irrespective of the degree of villous damage.[22,34,35] Furthermore, the seropositive cases with clear mucosal inflammation but normal villous morphology showed a beneficial response to GFD in most of the measured parameters. Nevertheless, it still remains unclear if all EmA-positive, screen-detected individuals should be treated. None of the participants in this study or in our previous randomized study investigating the role of EmA in diagnostics[15] had a completely normal biopsy (Marsh 0) at baseline and further prospective studies are needed to decipher whether these subjects truly benefit from a GFD. If so, even the universal need for a biopsy in EmA-positive adults might be questioned. Such studies, however, may be difficult to execute because a very large at-risk cohort likely should be screened to achieve sufficient power.

There were certain limitations to the present study. It was impossible to blind subjects to a GFD and some placebo effect may have accounted for the results. Furthermore, subjects in the gluten group might have been more sensitive to possible celiac disease–associated symptoms, and we cannot exclude gluten reduction at the individual level. Because EmA positivity was an inclusion criterion, there were no EmA-negative TG2ab-positive, or entirely seronegative, patients, and the results from this study should not be extended to apply to such individuals. Another limitation was that no other indicators for bone metabolism apart from BMD, for example, vitamin D levels, were measured. It also must be borne in mind that, in general, there is good knowledge of celiac disease and easy availability of gluten-free products in Finland, but this might not be the case in many other countries, for example, in the United States.[9,36] In contrast to celiac disease in general,[1] in this study 65% of the randomized subjects were males, the reason being that a considerably higher proportion of females than males fulfilled the age exclusion criteria.

In conclusion, our randomized study showed that apparently asymptomatic EmA-positive subjects benefit from serologic screening and a subsequent GFD. The results support active screening of celiac disease in at-risk groups. However, because of the possible detrimental effects of a GFD in social functioning, the consequences of screening must be prejudged on an individual basis. Furthermore, prospective studies are needed to unravel whether screen-detected seropositive subjects with completely normal small-bowel mucosal histology should be treated.