*Department of Molecular Medicine, University of Padua, Padua, Italy; †School of Specialization in Hygiene and Preventive Medicine, University of Padua, Padua, Italy; ‡Epidemiological Service, Udine, Italy; §Department of Cardiological, Thoracic and Vascular Sciences, University of Padua, Padua, Italy; ‖Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden; ¶Department of Pediatrics, Örebro University Hospital, Örebro University, Örebro, Sweden; **Division of Epidemiology and Public Health, School of Medicine, University of Nottingham, Nottingham, United Kingdom; and ††Department of Medicine, College of Physicians and Surgeons, Columbia University, New York, New York.
Inflamm Bowel Dis. 2017 Nov;23(11):1996-2000. doi: 10.1097/MIB.0000000000001098.
Celiac disease (CD) has been linked to inflammatory bowel disease (IBD) but previous reports have been inconsistent and may have been affected by surveillance bias.
Matched birth cohort study in Friuli-Venezia Giulia Region, Italy. We identified 1294 individuals with CD aged 0 to 23 years at diagnosis using pathology reports, hospital discharge records, or copayment exemptions. Each CD individual was matched with up to 5 general population reference individuals from the regional Medical Birth Register in Friuli-Venezia Giulia (n = 5681). As secondary comparison groups, we used individuals undergoing small intestinal biopsy but not having villous atrophy (either Marsh 0-1-2 or exclusively Marsh 0). Individuals with IBD were identified through hospital discharge records or copayment exemptions. Conditional logistic regression was used to estimate odds ratios (ORs) for having IBD among CD individuals (before or after CD diagnosis) compared with their matched references.
Overall 35 individuals with IBD were identified (29 with CD and 6 general population controls). This corresponded to an increased risk of IBD in CD (OR = 24.17; 95% CI, 10.03-58.21). However, compared with individuals with Marsh 0-1-2 the OR decreased to 1.41 (95% CI, 0.91-2.18) and restricting our comparison group to individuals with Marsh 0, the OR was 1.28 (95% CI, 0.61-2.70).
In conclusion, this article found a highly increased risk of IBD in individuals with CD when comparing with the general population. Bias is the likely explanation for the very high risk increase for IBD in CD because the excess risk was substantially lower when we used individuals with a small intestinal biopsy without villous atrophy as our reference.
乳糜泻（CD）与炎症性肠病（IBD）有关，但先前的报告不一致，并且可能受到监测偏差的影响。
意大利弗留利-威尼斯朱利亚地区的匹配出生队列研究。我们使用病理学报告、住院记录或共付额豁免，确定了 1294 名诊断为 0 至 23 岁的 CD 患者。每位 CD 患者都与弗留利-威尼斯朱利亚地区区域医疗出生登记处的至多 5 名普通人群参考个体进行匹配（n=5681）。作为次要比较组，我们使用接受小肠活检但无绒毛萎缩的个体（无论是 Marsh 0-1-2 还是仅 Marsh 0）。通过住院记录或共付额豁免确定 IBD 患者。使用条件逻辑回归估计 CD 患者（在 CD 诊断之前或之后）与匹配参考个体相比患有 IBD 的比值比（OR）。
总共确定了 35 名 IBD 患者（29 名 CD 和 6 名普通人群对照）。这对应于 CD 中 IBD 的风险增加（OR=24.17；95%CI，10.03-58.21）。然而，与 Marsh 0-1-2 的个体相比，OR 降低至 1.41（95%CI，0.91-2.18），并且将我们的比较组限制为 Marsh 0 的个体，OR 为 1.28（95%CI，0.61-2.70）。
总之，本文发现与普通人群相比，CD 患者的 IBD 风险显著增加。当我们使用没有绒毛萎缩的小肠活检个体作为参考时，IBD 的额外风险大大降低，因此，IBD 在 CD 中的风险增加过高可能是偏倚的解释。
