Introduction

The natural history of coeliac disease (CD) is still unknown. Clinically, CD appears both in childhood and adulthood, but whether CD can be acquired in adulthood is unknown. In addition, spontaneous normalization in serological tests has been observed in adults with initially positive serology. The aim of the present study was to assess serological signs of CD in pre-diagnostic samples of individuals from the general population with a subsequent CD diagnosis.

Aims & Methods

The study was based on the Trøndelag Health Study (HUNT). The HUNT study has been conducted in four comprehensive surveys inviting all adult inhabitants (>20 years of age) of Nord-Trøndelag County, Norway: HUNT1 (1984-1986), HUNT2 (1995-1997), HUNT3 (2006-2008) and HUNT4 (2017-2019). Blood samples were drawn from the participants during HUNT2, 3 and 4, and sera were aliquoted and stored at -80° Celsius at HUNT Biobank for future analyses. Of about 56,000 participants in HUNT4, 26,000 had participated in HUNT2 and 34,000 had participated in HUNT3, corresponding to approximately 20 and 10 years of follow-up, respectively.
Sera from participants in HUNT4 have were analysed for CD using a newly developed serological assay for detection of anti- transglutaminase 2 (anti-TG2) IgA. All seropositive participants were invited to a clinical validation study with upper endoscopy and tissue sampling to confirm the diagnosis histologically using the Marsh criteria. Available historical sera from HUNT2 and HUNT3 from seropositive participants in HUNT4 were analysed with the same serological assay.

Results

Of 686 seropositive participants in HUNT4 with available historical samples, 564 serum samples collected in HUNT2 and 594 serum samples collected in HUNT3 were analysed. Of the seropositive participants in HUNT4, 38% (215/564) and 24% (143/594) were seronegative in HUNT2 and HUNT3, respectively. Of 472 seropositive participants in HUNT4 with available historical sera from both HUNT2 and HUNT3, 19% (91/472) were seronegative in both surveys and 22% (104/472) were seronegative in one of the surveys. Hereof, 87 were negative in HUNT2, but positive in HUNT3. The remaining 17 were positive in HUNT2, but negative in HUNT3.
Of the 686 seropositive participants in HUNT4, 438 individuals have participated in the clinical validation study by March 2022. Of these, 62% (272/438) received a subsequent CD diagnosis and 8% (34/438) had already a biopsy-verified CD diagnosed before HUNT4.
Of 267 previously seronegative participants in HUNT2 and/or HUNT3 that participated in the clinical validation study, 3% (8/267) received a CD diagnosis before HUNT4 and 52% (139/267) received a subsequent CD diagnosis after HUNT4. Of 17 seropositive participants in HUNT2 with negative seroconversion in HUNT3, 2 participants received a subsequent CD diagnosis after HUNT4.

Conclusion

Our study demonstrates that CD may occur at any age, also in adulthood. In addition, a significant proportion of CD individuals may have lived with undiagnosed CD for several decades. Some individuals with negative serology, even with high IgA TG2 titres, can test negative later. Individuals with previous elevation of CD serology and conversion to negative serology are still be at risk of a subsequent CD diagnosis.